Papers

Jiale Amber Wang, Kaiyuan Wang, Pengyu Nie

Existing test generation benchmarks don't verify if tests actually run or match code changes; this benchmark solves that by grounding evaluation in real executable environments and commit history, revealing that state-of-the-art agents still struggle on recent tasks.

TestEvo-Bench is a benchmark for evaluating AI agents on test and code co-evolution tasks—writing new tests for code changes and updating failing tests. Unlike static benchmarks, it uses real commits from Java projects with executable environments to measure pass rates, coverage, and mutation scores.

Caleb Ziems, William Held, Su Doga Karaca et al.

Larger LLMs will simulate most human behaviors and opinions better, but scaling alone won't fix simulations of cognitive biases, rare populations, or tasks requiring specialized human knowledge—these need targeted research beyond just bigger models.

This paper investigates whether scaling up language models improves their ability to simulate human social behavior and opinions.

Achint Mehta

For agentic code generation, invest in reasoning capability and effort rather than external tools—stronger models and higher reasoning settings prevent failures at their root, while testing tools don't catch the reasoning errors that actually cause failures.

This study evaluated 90 runs of an agentic coding assistant building the same application, testing whether extra tools and prompts improve code quality. Results show that increased reasoning effort (not testing tools) dramatically improved first-try reliability, raising perfect runs from 28% to 89%, while a testing tool added 42-68% cost with no functional benefit.

Manuel Alonso-Carracedo, Ruben Fernandez-Boullon, Pedro Celard et al.

LLMs can grade technical exams reliably for simpler tasks, but struggle with complex questions—rubric quality matters more than which model you choose, and a taxonomy-based approach helps identify which questions are safe to auto-grade.

This paper evaluates whether large language models can reliably grade Linux/bash exam responses by testing GPT, Claude, Gemini, and GLM against expert instructors' grades.

M. Doris, S. Guo, S. M. Koh et al.

This package makes it practical for physics researchers to apply modern ML techniques (classification, object detection) to quantum gas experiments without building infrastructure from scratch.

Q-GAIN is a Python package that combines machine learning with physics-informed analysis for cold-atom experiments. It provides pre-built tools for classifying images, detecting objects, and analyzing quantum gas systems like Bose-Einstein condensates, with a modular workflow that connects data loading, ML-based feature detection, and physics analysis.

Thomas Winninger

Constraining coding agents through simple, language-level restrictions is more effective and cheaper than complex oversight frameworks—the same techniques that manage human engineering teams work better for AI agents.

This paper shows that applying traditional software engineering practices—access control, network policies, and coding conventions—to AI coding agents makes them safer and easier to oversee than using complex scaffolding.

Minghan Yu, Youran Sun, Chugang Yi et al.

Combining supervised retrieval scoring with zero-shot LLM reasoning can dramatically improve dataset discovery—achieving 5x better recall through score fusion and an additional 28% improvement through agentic reranking without extra training.

This paper presents an agentic search system that helps geoscience researchers find relevant datasets and tools from NASA's Earth Observation Knowledge Graph using natural language queries. The system combines supervised neural retrieval with LLM-based reasoning to significantly improve search accuracy, and includes a new benchmark of 47k query-dataset pairs for evaluation.

Benjamin Nichols, Michael Schlichtkrull, Nedjma Ousidhoum

MEDIAREF enables reproducible, cost-effective evaluation of how well LLMs can assess source credibility for fact-checking, addressing a gap where existing systems assume all retrieved evidence is equally reliable.

This paper introduces MEDIAREF, a public database of web documents from 200 media sources designed to help AI systems verify information credibility. Instead of relying on expensive search APIs, researchers can now use MEDIAREF to test how well language models assess whether news sources are trustworthy—a key step in fact-checking systems that cite their sources.

Lourdes Moreno, Paloma Martínez, Marco Antonio Sanchez-Escudero et al.

Multi-agent systems with internal quality signals and intelligent routing can outperform single-model approaches for text simplification tasks, even when adding lexical resources doesn't improve automatic metrics.

This paper presents three automatic systems for Spanish Easy-to-Read translation, submitted to a shared task. The best approach uses a multi-agent workflow combining two language models with quality signals and intelligent routing to simplify text while maintaining meaning. Results show this guided multi-agent approach outperforms simpler baseline methods.

Cristian-Gabriel Florea, Stelian Spînu

By running six specialized deep learning models locally on a smartphone with INT8 quantization, VisionAId achieves real-time visual assistance for blind and low-vision users without cloud dependency, while a few-shot learning pipeline lets users teach the system to find their personal objects.

VisionAId is an Android app that helps visually impaired people navigate and interact with their environment using on-device AI models. It combines depth estimation, object detection, and facial recognition to identify obstacles, locate personal items, and recognize faces—all running locally on a smartphone without requiring cloud connectivity, with optional AI for scene descriptions.

Zhi Chen, Zhensu Sun, Yuling Shi et al.

Performance-optimization benchmarks for coding agents have significant reliability issues: reference patches are unstable across machines, scoring rules heavily influence rankings, and most tasks are already solved by existing submissions, making leaderboard positions unreliable indicators of tru...

This paper audits three major benchmarks for evaluating coding agents on performance optimization tasks.

Austin McDannald, Julia Tisaranni, Howie Joress

Constraint programming can optimize experiment scheduling in autonomous labs by finding time-minimal plans that respect hardware constraints, with status dependencies ensuring robust real-world execution.

This paper tackles how to efficiently schedule experiments in autonomous labs when multiple instruments have different speeds and capacities. The authors use constraint programming to find optimal schedules that minimize total time while respecting hardware limits, then add a dependency system to ensure reliable execution of those schedules.

Sameer Malik, Ayush Singh, Amar Prakash Azad

By separating policy rules from document interpretation, PolicyGuard makes compliance review auditable and maintainable—you can inspect why a document failed review and update policies without retraining models.

PolicyGuard is a neuro-symbolic system that converts organizational policies into executable compliance review engines. It combines LLM-based document analysis with formal logic rules to check whether contracts and documents comply with company policies, making compliance decisions transparent and testable.

Yen-Jen Wang, Jiaman Li, Sirui Chen et al.

Synthetic data from reconstructed 3D scenes can effectively train perception-based humanoid robots for real-world loco-manipulation, eliminating the need for expensive human-annotated robot trajectories.

This paper solves a key bottleneck in training humanoid robots: the lack of paired data combining egocentric camera views, language instructions, and robot motion. The authors generate 48,000 synthetic training examples by reconstructing real indoor scenes with 3D Gaussian Splatting, simulating robot trajectories, and rendering first-person views.

Daniel Russo

Evaluating AI coding agents one at a time on isolated tasks misses the real problem: agent contributions create twice as much integration friction in shared codebases, making ecosystem-level governance more important than agent-level performance.

This paper studies how autonomous coding agents affect shared software repositories by analyzing over 930,000 pull requests. It finds that integration friction—the cost of merging code when others are changing it simultaneously—is largely a repository-level problem, not an agent problem.

Danyal Rehman, Charlie B. Tan, Yoshua Bengio et al.

Autoregressive models can outperform flow-based approaches for molecular sampling by avoiding invertibility constraints and enabling better scalability—opening a new direction for physics-informed generative modeling.

This paper introduces Autoregressive Boltzmann Generators (ArBG), a new method for efficiently sampling molecular systems at equilibrium. Unlike previous approaches using normalizing flows, ArBG uses autoregressive models to generate molecular configurations faster and more accurately, with a large pre-trained model (Robin) achieving 60% better energy predictions on peptide systems.

Kirill Solovev, Jana Lasser

Open-weight multilingual NLP can scale political network analysis beyond manual coding, extracting signed relationships from news at scale while remaining reproducible and avoiding proprietary APIs.

This paper presents an open-source pipeline for automatically extracting political relationships from multilingual news articles. It combines named-entity recognition, entity linking to Wikidata, and a specialized model to build knowledge graphs of political networks—showing it can reconstruct party lifecycles and uncover patronage networks in Austria and Poland.

Mohammad Mehdi Hosseini, Mohammad H. Mahoor, Hiroko H. Dodge

Language-based digital twins can authentically mimic individual conversational behavior while simultaneously serving as cognitive health monitors—enabling continuous, personalized monitoring of cognitive decline without requiring frequent clinical visits.

This paper creates AI-powered digital twins of elderly people using language models to capture their unique conversational patterns and writing style. The system learns from real conversations to generate authentic responses and can predict cognitive health scores, offering a non-invasive way to monitor early signs of cognitive decline.

Serhii Hamotskyi, Akash Kumar Gautam, Christian Hänig

LLMs can replace rigid rule-based systems for document compliance verification, handling messy real-world text better than traditional NER while requiring no task-specific training data.

This paper applies large language models to automatically verify whether securities meet eligibility criteria for use as collateral at the German Central Bank. Instead of manually reading through complex, bilingual prospectuses, the system uses LLMs to extract, normalize, and interpret financial and legal information, achieving 91% precision while avoiding false acceptances.

Santosh Kapuria, Abhishek

Transfer learning from simulation to real-world data enables accurate structural health monitoring with minimal labeled experimental data—a practical pattern for deploying deep learning in engineering applications with expensive real-world measurements.

This paper presents a transfer learning framework combining physics-based simulations and deep learning to diagnose structural damage using guided waves. By pretraining on cheap simulated data and fine-tuning on limited real experimental data, the approach achieves high accuracy for damage detection in structures with piezoelectric sensors while reducing computational costs.

Muhammad Hassan, Ramazan Yener, Ece Gumusel et al.

AI healthcare chatbots frequently fail users in critical ways—unreliable access, confusing interfaces, and privacy risks—suggesting designers need to prioritize reliability and trust before expanding these tools into healthcare.

This study analyzes over 15,000 user reviews of AI healthcare chatbots to understand how they work in real-world health information seeking. The researchers found three main problem areas: access and reliability issues, poor user experience, and billing/support problems, with privacy concerns causing the most negative user experiences.

Ping Liu, Qianqi Shen, Jianqiang Shen et al.

When training models with RL and AI feedback, reward signal design is critical—robust reward shaping prevents exploitation better than algorithm choice, and rule-based corrections can fix systematic failures like verbatim copying.

This paper tackles a real-world problem in job search: generating portable queries that capture candidate qualifications without user-specific details. The authors use reinforcement learning with AI feedback (RLAIF) to train models, but discover that standard reward signals get exploited—models learn to copy text verbatim instead of generalizing.

Maria Levchenko

Historical text imposes a consistent encoding penalty on LLMs, but models retain semantic understanding—making them safe for retrieval tasks if generative applications are adapted with temporal context.

This paper diagnoses why language models struggle with historical text by breaking down the problem into four dimensions: tokenization cost, predictive uncertainty, semantic robustness, and context sensitivity.

Parmitha Vangapandu, Sai Ganesh Mokkapati, Sathwik Narkedimilli et al.

Mental health NLP models perform better when trained on data that includes relational context and interpersonal triggers, not just isolated symptoms—this shift from individual-centric to context-aware modeling improves both accuracy and clinical relevance.

This paper introduces RSPC, a dataset of 1,799 Reddit posts about long-distance relationships annotated by psychiatrists for mental health conditions, relationship stressors, and relationship phases.

Chi Cui, Yixin Wu, Yang Zhang

AI nudification has democratized from targeting celebrities to harming people in users' social circles, enabled by accessible open-source tools and a tight-knit producer community that actively shares techniques.

This study analyzes 24,105 AI-generated non-consensual sexual images found on 4chan, revealing that targets have shifted from celebrities (4.7%) to everyday people (55.8%). Open-source models like Stable Diffusion power most production, while a small group of prolific creators drives the ecosystem through shared tutorials and fine-tuned models.

Fabiana Fournier, Lior Limonad

You can add agentic AI capabilities to existing business processes by wrapping them with policy-governed agents at specific control points, rather than rebuilding the entire system.

This paper introduces a 'process harness'—a layer that adds AI reasoning to existing business workflows without replacing them. It uses three types of AI agents (for tasks, decisions, and flow control) that operate within policy guardrails, letting legacy systems maintain structural control while gaining adaptive intelligence at key decision points.

Sen Li, Haichao Cui, Chendong Shao et al.

Unsupervised domain adaptation can transfer welding quality models between fundamentally different processes (laser vs. arc welding) without retraining from scratch, reducing the cost of deploying AI to new manufacturing equipment.

This paper solves a real manufacturing problem: deep learning models trained on one welding process fail when applied to another because the physical mechanisms differ. The authors use unsupervised domain adaptation to learn features that work across both laser and TIG welding, achieving 80%+ accuracy in cross-process transfer without needing labeled data from the new process.

Yves Ferstler, Adam Podoxin, Ty Brassington et al.

AI literary translation is competent but lacks the immersive quality readers value—and automatic metrics don't capture what makes human translation better, suggesting we need reader-centered evaluation for creative content.

Researchers compared AI-generated literary translations to human translations by having 15 avid readers evaluate excerpts from 15 recent novels in French, Polish, and Japanese. While readers found AI translations adequate, they consistently preferred human translations for their clarity and immersive quality.

Blade Frisch, Will Wade, Dylan Gaines et al.

AI for accessibility requires evaluation methods that go beyond standard metrics to capture the complex, intersectional needs of real users—one-size-fits-all approaches will fail people who depend on these systems.

This paper examines how AI can improve augmentative and alternative communication (AAC) systems for people with speech disabilities, but argues that standard evaluation metrics miss important nuances about what users actually need. The authors study six real AAC challenges and propose evaluation methods that account for the diverse, intersecting needs of individual users.

Ali Pourghasemi Fatideh, Wilder Baldwin, Maria Dhakal et al.

LLM dialogue systems for compliance assessment suffer from low accuracy against expert ground truth, and user satisfaction decreases with longer responses—designers should prioritize concise, proactive interactions over verbose explanations.

This paper evaluates how well LLM dialogue assistants (like GitHub Copilot) help developers assess non-functional requirements like HIPAA compliance. The researchers had 49 programmers use Copilot to evaluate 148 compliance requirements against real code, measuring both accuracy against expert standards and user satisfaction across multi-turn conversations.

Zidu Liu, Florian Marquardt

LLMs can solve discrete design problems in quantum computing by evolving structured concepts rather than generating solutions from first principles—showing that domain-specific constraints and executable specifications make AI search more effective.

Researchers used large language models paired with structured algebraic rules to automatically discover new quantum error-correcting codes. Instead of designing codes from scratch, the system evolves mathematical specifications and programs that describe code families, finding competitive designs that work better than some existing approaches.

Sara Dorfman, Maya Vishnevsky, Omer Dahary et al.

By generating diverse prompts rather than diverse images from one prompt, you can create navigable design spaces where each variation is semantically meaningful and user-understandable, rather than random visual differences.

This paper solves the diversity problem in text-to-image generation by shifting variation from the model's random sampling to the text prompt itself.

Sunil Wanjari, Manish Thakre, Aayushi Asole et al.

Combining multiple validated mental health screening tools with cognitive and personality data through a modular, interpretable system improves diagnostic accuracy and provides clinicians with more reliable decision support than using individual assessments alone.

PsyBridge is a decision-support system that combines clinical screening tools (PHQ-9, GAD-7), cognitive tests, and personality assessments into one interpretable framework for mental health evaluation.

Jincheng Zhong, Weizhi Wang, Che Jiang et al.

Enterprise agents need evaluation frameworks that measure artifact quality, cost, runtime, and skill transfer—not just task completion—because real workplace tasks are complex, heterogeneous, and require reproducible, auditable results.

This paper introduces EnterpriseClawBench, a benchmark for evaluating AI agents in real workplace environments. Built from actual enterprise sessions, it contains 852 reproducible tasks involving file handling, tool use, and business artifact creation.

Hengji Zhou, Ye Liu, Yufeng Liu et al.

You can generate personalized multimodal content on-demand by converting user behavior into preference signals, then using those signals to guide generation—avoiding the need to wait for or retrieve matching user-created content.

TailorMind generates personalized multimodal content (text + images) tailored to individual users by learning from their behavior patterns, even when no existing content matches their preferences. It combines collaborative filtering to understand user tastes with controllable generation to create novel, aesthetically pleasing content that stays true to authentic user patterns.

Campbell Lund, Thomas Euyang, Zanele Munyikwa et al.

Popular AI exposure scores are useful but incomplete—they don't capture how AI adoption actually happens, who really benefits or loses, or how impacts change over time.

This paper examines AI exposure scores—metrics measuring how much AI can assist with job tasks—that have become central to workforce policy debates. The authors show these static scores have significant limitations when applied to real-world policy questions, and identify a critical gap: researchers keep using outdated scores while ignoring newer methodological improvements.

Zahra Asghari Varzaneh, Reza Khoshkangini, Pia Saldeen et al.

Structured environmental data with domain-specific feature engineering can transfer across medical settings and improve clinical predictions—suggesting that underutilized sensor data is a valuable signal for healthcare AI.

This paper shows that detailed laboratory environmental monitoring—not just temperature and humidity averages—can meaningfully predict IVF pregnancy rates.

Bercan Turkmen, Vyas Raina

Prompting LLMs with multiple decompiler views of the same binary improves malware detection recall—a practical, training-free improvement for security analysts using AI to triage suspicious code.

This paper shows that using multiple decompilers (Ghidra and RetDec) to analyze the same malware binary improves LLM-based malware classification. Since different decompilers expose different artifacts through their lossy conversion process, combining their outputs helps LLMs better identify malicious code without requiring model retraining.

Yupeng Xie, Chen Ma, Zhenyang Wang et al.

Instead of manually creating data visualizations or using pixel-level video generation, DataMagic separates data logic from rendering through a declarative spec, ensuring videos accurately represent underlying data while supporting interactive exploration.

DataMagic is a system that automatically transforms raw tabular data and natural language questions into narrative data videos with charts, voice narration, and animations. It uses a declarative specification to ensure data accuracy and a multi-agent architecture to generate and organize video scenes, making data insights more engaging and explorable than static dashboards.

Denis Peskoff, Joe Barrow, Christopher Vu et al.

Local ordinances—which regulate zoning, housing, and public health—are now accessible as structured data for AI research, opening new possibilities for studying how everyday regulations are written and enforced across America.

LOCUS is a machine-readable corpus of nearly all U.S. local ordinances (codes from 9,239 cities and counties) that were previously locked in vendor platforms. The researchers used OCR to extract these fragmented documents and built an access layer covering 2,309 counties, enabling large-scale AI research on local law for the first time.

Michael Finkelson, Daniel Segal, Eitan Richardson et al.

You can generate multi-speaker audio with natural conversational qualities by using reference voices + free-form text prompts on foundation models, rather than building speech-only systems with rigid turn-based structure.

ScenA generates realistic multi-speaker audio scenes by conditioning a text-to-audio model on reference voice samples and natural language descriptions. Unlike systems that require structured turn-by-turn labels, it produces conversational audio with overlapping speech, background noise, and emotional sounds—mimicking real-world conversations.

Anoushka Vyas, Aarushi Dhanuka, Sina Khoshfetrat Pakazad et al.

Autonomous coding agents that execute and validate their own code outputs, combined with shared memory for experience reuse, outperform text-only approaches for data integration tasks and generalize across different SQL dialects and query types.

Data Intelligence Agents (DIA) is a production system using three autonomous coding agents to automate enterprise data workflows—interpreting raw data, creating schemas, and generating SQL queries.

Jiaqing Zhang, Sabyasachi Bandyopadhyay, Miguel Contreras et al.

Ambient sound and light sensors can be practical, non-invasive additions to ICU monitoring systems that meaningfully improve delirium risk prediction, with sound being the strongest signal.

This study uses ambient sound and light sensors in ICUs to predict which patients will develop delirium (acute confusion). Researchers tested neural network models on data from 309 patients across 9 hospitals, finding that sound levels alone predicted delirium with 80% accuracy, and combining sound with light improved short-term predictions.

Biswadeep Sen, Yi-Chieh Lee

Social chatbots should correct their own errors rather than outsource corrections to external sources, because self-correction preserves user trust and leverages the social relationship to amplify belief change.

When social chatbots make mistakes, how they fix them matters for user trust. This study tested three error correction approaches: external webpages, self-correction, and expert chatbots. Self-correcting chatbots maintained credibility better, and users who felt socially connected to the chatbot were more likely to believe the correction—but only when the chatbot corrected itself.

Shanda Li, Qiuhong Anna Wei, Jingwu Tang et al.

LLM agents can effectively identify reproducibility issues in ML papers by analyzing code repositories, even without execution—but they struggle with precise localization of problems.

ReproRepo is a scalable framework that uses GitHub issues as natural training data to evaluate whether AI agents can identify reproducibility problems in machine learning research. Testing on 1,149 papers, the best agent found at least one relevant issue for ~90% of papers, showing LLMs can spot real-world blockers without running code.

Joy Bose

This resource enables direct computational comparison of how independent interpretive traditions read the same source material—a capability previously unavailable at scale—and demonstrates both the value and limitations of stylometric and LLM-based approaches for analyzing philosophical disag...

Darshana Graph is a corpus of 125,000+ texts from Hindu, Buddhist, and Jain philosophical traditions, with a unique subset of 8,500 records where the same source verse is aligned across 18 historical commentators.

Diaa Fayed, Laurent Romary

A standardized, reproducible method for converting print dictionaries into machine-readable lexical resources—critical infrastructure for Arabic NLP that didn't exist before.

This paper describes how to convert a printed Arabic-English dictionary into a digital, machine-readable format using international standards (ISO LMF and TEI Lex-0). The authors tested their approach on a sample section and achieved 91% accuracy in parsing the dictionary's structure, while extracting synonyms and word features with 85-98% precision.

Weizhi Zhang, Zechen Li, Hamid Palangi et al.

For healthcare AI deployment, you need evaluation that's both scalable and clinically trustworthy—RubricsTree achieves this by using context-aware medical rubrics instead of relying solely on expensive human review or unreliable AI judges.

RubricsTree is a scalable evaluation framework for personal health AI agents that combines expert medical knowledge with automated assessment. It uses a hierarchical taxonomy of 100+ clinically-verified rules that adapt to each query, solving the problem of expensive physician review while avoiding the inconsistency of AI judges.

Jiaju Han, Ben Zhang, Xuemeng Sun et al.

Combining infrared and RGB satellite imagery with modality-specific captions significantly improves vision-language models' ability to understand Earth observation data—infrared-aware text supervision is essential for effective multi-modal learning.

FusionRS is the first large-scale dataset pairing RGB and infrared satellite images with text descriptions for training AI models that understand both types of imagery together.

Gaurav Verma, Scott Counts

LLMs can reliably abstract noisy, granular user interaction logs into interpretable workflows without domain-specific training, enabling better product insights while preserving privacy.

WorkflowView uses large language models to convert low-level user action logs (clicks, keystrokes, etc.) into high-level, meaningful activities. The approach works across different applications and domains—from browser history to online courses to document editing—without needing task-specific training, making it practical for understanding real user behavior at scale.

Tobias Holtdirk, Pietro Marcolongo, Anna Steinberg Schulten et al.

LLMs can automate reproducibility assessment in social sciences at scale, matching or exceeding human reanalysts' ability to verify whether published findings hold up when the data is reanalyzed.

Researchers tested whether large language models can automatically check if published social science studies are reproducible by having LLMs reanalyze data and compare results to original findings.

Amy Xin, Jiening Siow, Junjie Wang et al.

The key bottleneck for autonomous scientific discovery isn't better agent algorithms—it's better environment design. By engineering constraints, collaboration tools, and oversight mechanisms, you can make agents more reliable and productive at discovering new solutions.

EurekAgent is an AI agent system designed to automate scientific discovery by focusing on environment engineering—the resources, constraints, and interfaces that shape agent behavior.

Kyuil Lee, Dezhi Yu, Yongkang Huang

For symbolic music generation, autoregressive models with attention outperform VAEs and GANs at producing stylistically coherent Bach compositions, though vector quantization significantly improves VAE performance by preventing posterior collapse.

This paper compares three types of AI models for generating Bach-style piano music from symbolic notation: autoregressive models (which predict one note at a time), latent-variable models (which learn compressed representations), and adversarial models (which compete to fool each other).

Minghao Luo, Liang Chen

Search-augmented LLMs used for recommendations are highly vulnerable to web content manipulation; even a single fake review or promotional page can mislead models into promoting non-existent products, and common defenses like reasoning and skepticism prompting often backfire.

This paper tests how easily AI recommendation systems can be tricked by fake product reviews and promotional content planted on websites. Researchers created FORGE, a benchmark that simulates web pollution by replacing real products with fake ones in search results, then measured how often 12 different LLMs recommend the fake products.

Haochen Wu, Yi Hou, Shiguang Xie

You can safely deploy RL in production by having learned policies adjust existing systems rather than replace them, using offline learning from delayed marketplace feedback with conservative value estimation to avoid overoptimistic decisions.

DoorDash built a reinforcement learning system that learns to adjust how their delivery dispatch algorithm balances speed vs. efficiency using real marketplace feedback. Instead of replacing the core optimizer, a learned policy selects adjustment multipliers based on delayed signals like delivery times and courier workload, enabling safe offline learning from noisy production data.

Harihara Muralidharan, Reema Baskar, Soo Hee Lee et al.

Current AI agents fail at domain-specific scientific reasoning in genomics: they can locate data and perform calculations, but lack the deeper understanding needed to make correct analytical decisions for specialized assays.

EpiBench is a benchmark that tests whether AI agents can perform epigenomics analysis tasks—like analyzing DNA sequencing data from CUT&Tag, ATAC-seq, and ChIP-seq experiments—by making correct decisions and returning verifiable answers.

Tanmoy Kanti Halder, Akash Ghosh, Subhadip Baidya et al.

Specialized medical AI systems need language-specific and multimodal training to serve non-English-speaking populations effectively—this work shows how to build such systems with structured reasoning agents.

ArogyaSutra addresses healthcare AI gaps in rural India by introducing a multilingual medical dataset covering 21 clinical domains in English and seven Indian languages, plus a multi-agent framework that reasons through medical questions step-by-step using tool integration and memory mechanisms.

Guojun Liao

AI's biggest gap in scientific discovery isn't search or execution—it's the ability to recognize structural inadequacy in existing frameworks and find solutions through conceptual insight rather than brute-force optimization.

This paper argues that AI in scientific discovery requires three layers: searching existing knowledge, forming new models through structural insight, and executing solutions. The key innovation is Layer 2—recognizing when current frameworks are inadequate and finding solutions by understanding what's missing conceptually, not through trial-and-error.

Shengqiang Zhang, Ruotong Liao, Volker Tresp et al.

By working with a VAR model's native bitwise predictions and residual code composition instead of token streams, you can edit images more precisely while preserving backgrounds—no retraining needed.

BitResEdit is a training-free editor for visual autoregressive image generators that uses two techniques: BitEdit guides bit-level predictions toward text descriptions, while ResEdit applies edits through the model's native residual code structure. This approach preserves unedited regions exactly while making precise, localized changes to match text prompts.

Fabien Chraim, Jian Zhang, Dominik Janzing et al.

Self-supervised learning can model fault propagation in networks well enough to causally attribute customer impact to root causes, enabling faster incident response than traditional rule-based systems.

NetCause learns how faults spread through networks to identify root causes of outages. Instead of using fixed rules, it trains on real incidents to understand fault propagation patterns, then uses counterfactual simulation to rank which component most likely caused a customer impact. Tested on production cloud network data, it outperforms rule-based approaches by 16%.

Haotao Xie

Domain-specific datasets and targeted fine-tuning can significantly improve LLM performance on specialized tasks like classical poetry understanding, showing that treating niche domains as general problems leaves performance on the table.

This paper creates a specialized dataset of 49,404 instruction-response pairs for classical Chinese poetry tasks and fine-tunes Qwen2.5-14B using LoRA to build PoetryQwen. The model breaks poetry understanding into three subtasks—term interpretation, semantic interpretation, and emotional inference—and achieves 9.7% improvement over the baseline on a poetry evaluation benchmark.

Zhiyi Chen, Jie Song, Peng Li

You can significantly improve Text-to-SQL accuracy without retraining models by systematically capturing and reusing error patterns as retrievable hints—Tahoe achieved 79% pass rate (up from 62%) on Spider 2.0 by learning from just 113 examples.

Tahoe is a system that improves Text-to-SQL performance by learning from errors and building a reusable library of hints. Instead of retraining models, it captures compiler feedback and user corrections as structured hints that guide LLMs to generate correct SQL for specific databases and dialects.

Zhi Wei Xu, Torbjörn E. M. Nordling

Robots can now reliably measure human heart rate from RGB cameras in varying lighting by combining spatial-temporal transformers with illumination-aware training, enabling non-contact physiological monitoring for human-robot interaction.

This paper develops a camera-based system for robots to measure human heart rate from video, even when lighting conditions change dramatically.

Haoyuan Deng, Yitong Gao, Yudong Lin et al.

Autonomous intervention systems can dramatically reduce human workload in robot learning by predicting when policies are failing and self-correcting using learned recovery strategies from past experience.

UniIntervene is an AI system that learns to intervene in robot learning tasks without waiting for humans. Instead of requiring frequent human corrections, it detects when a robot is stuck in unproductive exploration, retrieves successful past solutions from memory, and executes corrective actions autonomously.

Kevin Qinghong Lin, Batu EI, Yuhong Shi et al.

Building trustworthy AI journalism requires end-to-end orchestration of specialized agents plus explicit evidence-grounding: every number and claim must link back to verifiable data, code, or references.

Data2Story is a multi-agent system that automates data journalism by transforming raw data into complete multimedia news stories. It combines specialized agents for data analysis, writing, and design while ensuring every claim is traceable back to its source—making stories more transparent and verifiable than traditional approaches.

Abhijoy Sarkar, Aarchi Singh Thakur

Building predictive models for cancer drug resistance requires longitudinal data (serial measurements over time), not just a single snapshot of tumor genetics; current single-timepoint data hits a hard ceiling regardless of algorithm sophistication.

OncoTraj is a public benchmark dataset of 813 lung cancer patients for predicting treatment resistance to osimertinib. It combines data from three clinical sources and defines three prediction tasks: whether patients progress within 12 months, how long until progression, and what resistance mechanism emerges.

Avijit Ghosh, Anka Reuel, Jenny Chim et al.

Inconsistent evaluation reporting across AI leaderboards and papers makes it impossible to reliably compare models—Evaluation Cards solves this by creating a unified, machine-readable format that surfaces what's missing and helps different stakeholders understand what results actually mean.

This paper introduces Evaluation Cards, a standardized reporting system that makes AI model evaluation results comparable and interpretable across different sources.

Matthew Ho, Brian Liu, Jixuan Chen et al.

Coding agents can operate complex scientific software efficiently by learning just the simulator's interface contract (vocabulary, constraints, validation rules) rather than the full domain—and these adapters can improve themselves by learning from past attempts.

SIGA helps general-purpose coding agents learn to use specialized scientific simulators by providing a lightweight adapter that teaches them the simulator's input language, validation rules, and constraints.

Lawrence Keunho Jang, Mareks Woodside, Geronimo Carom et al.

Current AI agents struggle with multi-app tasks (37% success) and personalization—they need access to user context and device state to act intelligently, not just follow isolated instructions in sterile environments.

iOSWorld is a benchmark for testing AI agents on real iOS phones with persistent user data. Unlike sandboxed tests, it evaluates whether agents can reason about a user's identity, history, and preferences across 26 interconnected apps with realistic data like messages, transactions, and travel records.

Jeremy Yang, Kate Zyskowski, Noah Yonack et al.

Autonomous AI agents don't just answer questions faster—they fundamentally change what work users attempt by automating task decomposition and execution, shifting knowledge workers toward higher-order activities like verification and strategy rather than manual orchestration.

This paper analyzes how autonomous AI agents reshape knowledge work by comparing Perplexity's Search (conversational) and Computer (agentic) products.

Chris R. Jung, Markus Dörr, Natalie Jüngling et al.

Drifting models can replace slow iterative diffusion for CFD surrogates, enabling real-time flow field generation that's orders of magnitude faster while matching diffusion model quality.

This paper speeds up CFD simulations by using a generative model called "drifting" instead of traditional diffusion models. The model learns to generate realistic fluid flow patterns in a single pass rather than iteratively, making it 100x faster while maintaining accuracy. It uses a learned latent space and label-aware masking to ensure generated flows match boundary conditions.

Jui-Hui Chung, Ziyang Cai, Zihao Li et al.

Blueprint-based theorem proving—planning the proof structure upfront before solving individual lemmas—is more efficient than recursive decomposition and enables solving hard competition math problems with open-source models at low cost.

Goedel-Architect is an AI system that proves formal mathematical theorems in Lean 4 by first generating a blueprint—a dependency graph of definitions and lemmas needed for the proof—then solving each lemma in parallel.

Gandhimathi Padmanaban, Fred Feng

For real-world computer vision deployment, a two-stage pipeline with confidence-based abstention (refusing to predict when uncertain) is more reliable than forcing predictions—the model's uncertainty directly predicted where it would fail on new data.

Researchers built an open-source tool that automatically identifies vehicle types from roadway video to help understand cyclist safety in crashes. The system uses two AI models in sequence: first detecting vehicles in video frames, then classifying them into six specific body types (car, SUV, pickup truck, etc.).

Linyao Chen, Qinlao Zhao, Zechen Li et al.

LLM agents can improve prediction accuracy on ambiguous cases by dynamically deciding what evidence to gather, rather than using static prompts—showing a practical way to combine LLM reasoning with structured decision-making.

This paper presents AgentMob, an LLM-based agent that predicts where people will go next by adaptively gathering evidence rather than relying on fixed prompts. It uses a fast path for routine movements and triggers iterative tool use (checking historical patterns, geography, movement likelihood) when predictions are uncertain, achieving strong results without task-specific training.

Zhifei Xie, Zihang Liu, Ze An et al.

Real-time audio AI is now possible with a single model that can listen, understand, and respond continuously—moving beyond offline systems that handle only one task at a time.

This paper introduces Audio-Interaction, a unified streaming model that listens to audio in real time and responds on the fly, rather than processing audio offline like current systems.

Roohan Ahmed Khan, Yasheerah Yaqoot, Muhammad Ahsan Mustafa et al.

Instead of manually designing reward functions for robot learning, use an AI agent to generate, evaluate, and refine rewards automatically—this reduces human effort and improves policy performance by 71% through closed-loop self-improvement.

AgenticRL is a framework that uses a multimodal AI agent to automatically design reward functions, train drone navigation policies, and refine them through feedback loops—eliminating manual reward engineering.

Máté Gedeon, Péter Mihajlik

LLM-generated synthetic conversations paired with TTS can effectively replace scarce real conversational data for training speech recognition systems, especially when real multi-speaker dialogue is expensive to collect.

This paper shows how to train better speech recognition systems for low-resource languages by generating fake conversations using LLMs and text-to-speech. Instead of collecting expensive real conversations, the authors create synthetic multi-speaker dialogues with realistic speaker metadata, then use TTS to generate audio.

Yuxing Lu, Yushuhong Lin, Wenqi Shi et al.

Current LLMs struggle with clinical decision-making not just in what they decide, but critically in how they gather information—they ask redundant questions and fail at management decisions even when they get diagnoses right, revealing a gap invisible to outcome-only evaluation.

ClinEnv is an interactive benchmark that tests how well AI models act as doctors by simulating real patient cases over multiple decision stages. Unlike static medical benchmarks, it requires models to actively gather information from specialized agents before making treatment decisions, and scores both the quality of decisions and the process of gathering information.

Liuliu Chen, Gowri Rajaram, Eleanor Bailey et al.

LLM-augmented ML can extract clinical signals from unstructured text (triage notes) for public health surveillance, achieving high accuracy while remaining transferable across hospitals—showing practical value for real-world healthcare deployment.

Researchers built a machine learning system to detect self-harm cases from emergency department triage notes by combining traditional ML with large language models. The system identifies self-harm incidents and the specific method used, and works reliably across different hospitals without needing retraining for each location.

Qinpei Luo, Ruichun Ma, Xinyu Zhang et al.

Representation design matters more than model size: a semantically-grounded code format for PCB schematics lets smaller LLMs outperform larger general-purpose models on hardware design tasks.

SchGen is the first AI system that generates PCB schematics from natural language descriptions. The key innovation is a new code-based representation that focuses on semantic relationships (like component connections) rather than geometry, making it easier for language models to learn. The authors also built a large dataset by converting open-source hardware designs into this representation.

Keshigeyan Chandrasegaran, Kyle Sargent, Suchir Agarwal et al.

For building image generation models, GPIC provides a legally usable, large-scale alternative to web-scraped datasets—eliminating licensing concerns while offering standardized evaluation benchmarks.

GPIC is a massive dataset of 100M training images (28 trillion pixels total) with AI-generated captions, all permissively licensed for research and commercial use. The dataset is deduplicated, safety-filtered, and hosted on Hugging Face with benchmarking tools and baseline models for training visual generation systems.

Ruixiang Jiang, Chang Wen Chen

Pre-capture planning for portraits is more effective than post-production editing: by reasoning about 3D scene geometry, lighting physics, and aesthetic principles before shooting, you can automatically generate better photography setups.

This paper tackles portrait photography planning before taking the photo—deciding where to position the subject, where to place the camera, and how to set up lighting in a 3D scene.

Chong Bao, Shichen Liu, Lijun Yu et al.

A single pretrained model can now generate realistic digital humans across all modalities (speech, movement, appearance) by treating them as a unified problem rather than separate tasks, making it practical to build avatar systems without specialized sub-models.

Archon is a unified AI model that generates digital humans across multiple modalities—text, audio, motion, and video—all from a single system. It uses a novel video compression technique to handle high-resolution talking videos efficiently, and a step-by-step reasoning approach that switches between modalities to improve output quality.

Valentina Bui Muti, Eugénie Dulout, Ziquan Fu

LLMs show significantly lower diagnostic accuracy when working with structured healthcare data formats (FHIR) compared to plain text, meaning benchmarks must match actual clinical system requirements to predict real-world performance.

This paper creates a dataset of realistic medical records in FHIR format (a standard healthcare data structure) to test how well AI models perform clinical reasoning.

A. J. Lew, Y. Cao, M. J. Buehler

LLMs show promise for scientific discovery tasks—GPT-4 variants maintain strong alignment with real research conclusions even with limited context—but current benchmarks don't adequately test the creative hypothesis generation needed for genuine scientific breakthroughs.

This paper introduces ProjectionBench, a benchmark that tests whether large language models can generate scientific hypotheses like real researchers do. Models receive a research question with details gradually revealed, and their generated hypotheses are compared to actual paper conclusions using semantic similarity.

Audrey Chan, Aaron Labbé, Jacob Lavoie et al.

When you can't safely experiment on users (especially vulnerable populations), build a world model from past data to simulate outcomes, then optimize your recommendation policy offline—this avoids ethical risks while still improving emotional outcomes.

This paper describes AMRS, a music recommendation system for clinical and wellness users that predicts how songs affect emotional states (valence and arousal) using a world model trained on listening data.

Rishi Bommasani, Sarah H. Bana, Kathleen A. Creel et al.

When many employers use the same hiring algorithm, it amplifies bias rather than spreading risk—the same people get rejected everywhere, and racial disparities compound across the job market.

This paper analyzes hiring algorithms from a single vendor used by many employers and finds they create unfair outcomes.