Papers

Xuanyu Chen, Nan Yang, Shuai Wang et al.

When training on decentralized, non-uniform data, use Masked Image Modeling instead of Contrastive Learning—it's theoretically more robust. Better network connectivity always improves robustness, so federated learning is a viable alternative to fully decentralized systems.

This paper analyzes how distributed self-supervised learning systems handle non-uniform data across devices. The researchers prove that Masked Image Modeling is more robust to data heterogeneity than Contrastive Learning, and that federated learning performs as well as fully decentralized approaches. They introduce MAR loss, a practical improvement that aligns local and global representations.

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.

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.

Nina Begus

Literary disciplines offer practical tools for making AI systems more culturally literate and pluralistic, moving beyond the monolingual, automated cultural encounters that current LLMs create.

This essay argues that literary analysis methods—comparative reading, narratology, critical theory, and world literature approaches—are essential for building culturally aware AI systems.

Chih-Han Yang, Dai-Jie Wu, Yun-Ping Huang et al.

Language-based feedback preserves more information than scalar signals when learning from imperfect data, enabling policies to understand not just what went wrong but why and how to fix it.

This paper proposes using natural language critiques as structured supervision signals for learning from suboptimal demonstrations. Instead of compressing feedback into scalar scores, the method generates language labels describing task progress, failures, and corrections, then trains policies directly on these rich signals.

Yuqing Yang, Qi Zhu, Zhen Han et al.

Data referencing errors are a widespread problem in LLM table reasoning that goes beyond final-answer accuracy; using a lightweight critic model to catch these errors during inference significantly improves reliability.

LLMs make data referencing errors when reading tables—citing wrong values or missing data despite understanding table structure. This paper systematically measures these errors across models and shows that using a critic model to detect and filter bad outputs improves accuracy by up to 12%, even with a small 4B-parameter critic.

Kaisen Yang, Zheng Jiang, Yuzhao Peng et al.

Browser agents can scale more efficiently by learning from the implicit skills already present in human web interactions rather than from manually designed tasks, using skill distillation to convert trajectories into reusable, composable natural-language instructions.

This paper proposes a scalable approach for training browser agents by distilling human web browsing interactions into reusable natural-language skills. Rather than training agents from scratch on individual tasks, the method converts user interaction traces into compact skill descriptions that agents can retrieve and compose, organized in a skill graph to enable efficient learning and reuse.

Ekaterina Alimaskina, Denis Shveykin, Gleb Molodtsov et al.

Synthetic QA generation for model training has hidden failure modes: biased coverage of documents and susceptibility to instruction injection. Simple fixes like anchoring questions to specific targets and filtering instruction-like text can substantially reduce these problems.

This paper reveals that using synthetic question-answer pairs to train language models is riskier than assumed. Models generating QA pairs don't uniformly cover documents—they focus on salient regions and can be hijacked by artifacts like markup.

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.

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.

Lang Huang, Jinglue Xu, Luke Darlow

Before building bigger models, optimize your data preprocessing: context length, normalization strategy, and regularization can close most of the accuracy gap at a fraction of the computational cost.

This paper shows that simple linear models (Ridge regression) can match or beat complex deep learning architectures for time-series forecasting by carefully tuning preprocessing—context length, normalization, and regularization—rather than scaling model size.

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.

MSVPJ Sathvik, Parmitha Vangapadu, Nishit Rane et al.

The dataset enables building explainable AI systems to automatically detect manipulative betting ads on social media, with practical applications for user protection and regulatory monitoring.

This paper introduces BetXplain, a dataset of betting advertisements from Instagram and Reddit annotated for manipulative tactics and deceptive practices. Each ad includes human explanations of why it's manipulative, enabling research into detecting misleading betting promotions that could harm users' mental health and financial well-being.

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.

John Sweeney

FisherSketch enables practical source selection for LLM families by measuring task similarity through Fisher alignment signatures (16 KB per task) instead of expensive full Fisher matrices, revealing whether tasks differ in activations, errors, or their interaction.

This paper solves the problem of selecting training data sources for language models that share vocabularies but differ in tasks (like SMILES vs protein sequences).

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.

Bo Chen

Statistically significant findings from keyword-based text analysis can be entirely artifacts of the measurement method rather than real phenomena. Always validate keyword-based results with semantic approaches before drawing conclusions about speaker psychology or discourse patterns.

This paper reveals how keyword-based measurement tools can produce false findings in computational social science. By comparing keyword counting to LLM-based semantic analysis of interviews, the authors show that a strong statistical correlation between negative affect and certainty disappears—and even reverses—when using more accurate measurement.

Juliana Li, Diya Sreedhar

Language models don't just forget rules randomly during training—their survival is determined by corpus statistics (support frequency), and this process is irreversible: you can kill learned behaviors but cannot resurrect them through data manipulation.

During language model pretraining, learned rules like pronoun-gender agreement mysteriously disappear mid-training even though evidence for them remains in the data. This 'natural ungrokking' is predictable: rules survive based on how often the training data supports them relative to competing patterns.

Negin Raoof, Richard Zhuang, Marianna Nezhurina et al.

Systematic data curation matters more than you might think—the right mix of task sources and diversity in training data significantly improves how well agents generalize across different benchmarks.

This paper presents OpenThoughts-Agent, an open framework for creating training data for AI agents that can handle diverse tasks. The authors ran 100+ experiments to understand what makes good training data, then created a 100K example dataset that improved agent performance by 3.9 percentage points over existing open models.

Maria Nefeli Paraskevopoulou, Tatiana Passali, Grigorios Tsoumakas

For scientific summarization, training on carefully selected high-quality examples outperforms training on larger random datasets—quality matters more than quantity when building summarization systems.

This paper creates a large biomedical summarization dataset (1.88M articles) and shows that author-written abstracts vary in quality. By selecting high-quality training examples based on alignment with source articles, models achieve better results with less data than random sampling, improving both efficiency and factuality.

Hariom Ingle, Ronit Ghode, Ishwari Gondkar et al.

Marathi NLP now has a gold-standard POS tagging dataset and baseline models, enabling better downstream tasks like machine translation and parsing for an under-resourced language with 83+ million speakers.

This paper introduces L3Cube-MahaPOS, a manually annotated dataset of 32,354 Marathi sentences for part-of-speech tagging, along with benchmarks across multiple model architectures.

Jinghong Lan, Wei Cheng, Yunuo Chen et al.

By mining community LoRAs as style-content anchors and using curriculum learning with targeted disentanglement mechanisms, you can scale dual-reference image generation while maintaining clean separation between style and content.

FreeStyle is a framework for generating images that combine the style of one image with the content of another. It uses community LoRA models as building blocks to create large training datasets and employs specialized techniques to prevent unwanted style or content leakage. The approach includes a new benchmark for evaluating dual-reference image generation.

Yusuf Salcan, Simon Ging, Robin Schirrmeister et al.

You can train medical vision-language models to perform spatial grounding (locating regions in images) alongside report generation without sacrificing language quality, using automatically-curated training data instead of expensive manual annotations.

This paper introduces RefRad2D, a large-scale bilingual dataset of 1.2M medical images paired with text, and RadGrounder, a vision-language model trained to simultaneously generate radiology reports, answer visual questions, and locate anatomical regions via bounding boxes or segmentation—all without manual spatial annotations.

Bruno Scarone, Alfredo Viola, Renée J. Miller

You can reduce bias in ML models by strategically modifying training data, but there's a trade-off: stricter fairness requirements cost more in data changes, and ensuring sufficient representation of intersectional groups is crucial for both fairness and model performance.

This paper addresses how to reduce bias in machine learning models, especially for underrepresented groups defined by multiple characteristics (like race and gender together). The authors propose a method that modifies training data to reduce bias while ensuring enough examples exist for all groups, and they measure the cost of achieving different levels of fairness.

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.

Zahra Abbasiantaeb, Zeno Belligoli, Omar Essam et al.

Style diversity in synthetic data is more important than topic diversity for training intent classifiers—varying how things are said matters more than varying what's discussed.

This paper presents a method for generating synthetic training data for intent classification without any human annotations. The approach uses intent definitions and LLM generation with style and topic diversity controls, plus post-hoc stylization models to create varied, realistic dialogue. Results show the synthetic data reaches 93% of the performance of human-annotated data.

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.

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.

Christopher B. Womack, Shahine Bouabid, Andrei Sokolov et al.

For climate emulators, optimizing training data diversity through iterative refinement produces better generalization than simply using more standard scenarios, even with smaller datasets.

This paper shows that training data quality matters more than quantity for climate AI models. Instead of using many standard climate scenarios, researchers created a method to design fewer but more diverse training scenarios that teach AI models to better predict climate behavior across different conditions—like distinguishing how greenhouse gases versus aerosols affect the climate.

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.

Trisha Mittal, Akshay Mehra, Joshua Kimball

Dataset distillation doesn't consistently outperform simpler coreset selection methods on real-world tasks, and coresets are often more computationally efficient for creating compact training datasets.

This paper challenges the effectiveness of dataset distillation by comparing it against simpler coreset selection methods across large-scale benchmarks.

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.

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.

Dimitri Kachler, Damien Sileo, Pascal Denis

You can dramatically speed up identifying problematic training data by training a lightweight encoder to predict influence rankings, rather than computing expensive influence functions directly.

This paper proposes Influcoder, a method that speeds up data attribution for large language models by distilling influence function computations into a compact encoder. Instead of expensive influence calculations, it learns to quickly identify which training samples contributed to specific model outputs, enabling efficient dataset curation and quality filtering.

Lezhi Tan, Tijana Zrnic

You can use synthetic data for research if you can prove your task is exchangeable with historical tasks where real data is available; this framework provides statistical guarantees that your conclusions remain valid.

This paper provides statistical methods for safely using synthetic data in research by introducing 'task exchangeability'—a condition ensuring your current research question is mathematically similar to past tasks where real data exists. The authors develop inference techniques with validity guarantees and test them on LLM-generated survey responses and AI evaluation tasks.

Henrique Ferraz de Arruda, Andreia Sofia Teixeira, Pranay Gundala Reddy et al.

Mental health conversations on TikTok show a gap between creator intent and audience response: videos express negative emotions about serious topics, but comments shift toward positivity, though toxicity concentrates in certain sensitive areas like suicide prevention.

Researchers analyzed 28,341 TikTok videos and 80,130 comments from Mental Health Awareness Month to understand how mental health content is framed and received.

Yifan Hu, Hongzhou Chen, Peiyuan Liu et al.

For practical time series forecasting with missing data, you need to predict both whether observations will happen and what their values are—treating these as a coupled problem rather than assuming future timestamps are known.

This paper tackles incomplete time series forecasting by jointly predicting whether future observations will occur and what their values will be. Most existing methods assume future timestamps are known, which isn't realistic.

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.

Noémi Éltető, Nathaniel D. Daw, Kimberly L. Stachenfeld et al.

Active learning can dramatically accelerate scientific discovery by automatically designing experiments that test competing mechanistic theories, reducing the data needed to understand complex systems.

ATLAS is an automated system that discovers interpretable behavioral models by iteratively generating competing hypotheses and designing experiments to distinguish between them. It uses sparse neural networks and active learning to efficiently recover how agents make decisions, achieving 5-10x better sample efficiency than random experimentation.

Sanjay Adhikesaven, Haoxiang Sun, Sewon Min

Modern LLMs depend on dozens of upstream models and datasets in complex, recursive ways; ModSleuth makes these invisible dependencies visible and traceable, exposing compliance and transparency issues in LLM development.

This paper introduces ModSleuth, a system that automatically traces the hidden dependencies between models and datasets used to build modern LLMs.

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.

Alexander Chulzhanov, Soeren Eberhardt, Arjun Mukherjee

Synthetic data can efficiently teach a model grammatical structure for low-resource languages, but semantic understanding requires authentic data—synthetic bootstrapping works best as a primer before curriculum learning with real examples.

This paper tackles machine translation for Q'eqchi' Mayan, an Indigenous language with almost no digital text. Instead of scraping the web (which violates data sovereignty), researchers created synthetic training data from dictionaries and fine-tuned a multilingual model using LoRA adapters.

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.

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.

Yalun Dai, Yangyu Huang, Tongshen Yang et al.

The sequence of training data is as important as which data you select; reordering data using simple principles can boost LLM training efficiency and stability with minimal overhead.

This paper shows how the order in which you feed data to language models during training matters significantly. The researchers identified four key principles for organizing training data—Boundary Sharpening, Cyclic Scheduling, Curriculum Continuity, and Local Diversity—and created two new data ordering methods that improve training stability and performance without extra computational cost.

David Busbib, Michael Werman

Generating mathematically sound future claims requires grounding in both scientific context (citations) and formal logical structure—neither alone is sufficient for plausible theorem prediction.

This paper tackles predicting future mathematical theorems by combining two types of information: citation networks (what mathematicians cite) and formal theorem dependencies (what logically follows from existing proofs). The authors build COMPOSE, a system that uses both graphs together to generate plausible new theorem statements, validated against real papers from 2024-2025.

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.

M. Ross Kunz, John Merickel, Keith Wilson

You can now retrieve similar datasets and understand which variables correspond across them without needing shared column names—useful for finding relevant training data or initializing models for new datasets.

This paper presents a method to compare and retrieve numeric datasets by converting their statistical properties into embeddings. Instead of requiring datasets to share the same variables, the approach uses statistical summaries (like means, distributions) embedded via sentence transformers, then applies Canonical Correlation Analysis to find which variables align across datasets.

Haowen Wang, Yaxin Du, Jian Yang et al.

For mid-training, let the data sources themselves tell you what quality matters: MIRA discovers source-specific rubrics automatically, making data selection both more effective and more scalable than fixed evaluation criteria.

MIRA is a data selection method for mid-training large language models that automatically discovers what quality criteria matter for each data source, then uses those criteria to filter training data efficiently.

Beiduo Chen, Pingjun Hong, Ziyun Zhang et al.

Instead of treating annotator disagreement as noise, you can use it as a signal to train models that reproduce individual annotators' reasoning styles—useful for building explanation systems grounded in real human decision-making patterns.

This paper shows that large language models can learn to mimic how individual annotators explain their decisions, not just their labels. The researchers developed a method called CAPO that trains models on one annotator's explanations while contrasting them with other annotators' valid but different explanations for the same input, revealing stable patterns in how people reason about tasks.

Shiyu Chen, Tarfah Alrashed, Alon Halevy et al.

For agents to reliably execute data-driven tasks, structured semantic metadata is essential—unstructured web search alone leads to frequent failures where agents retrieve descriptions instead of actionable data.

This paper compares how AI agents find data on the web: one searches billions of unstructured web pages, while another uses structured semantic metadata (schema.org). The semantic approach retrieves more usable data with 65.7% higher precision, while the unstructured approach covers more questions but often returns unhelpful pages instead of actual datasets.

Yi Jing, Zao Dai, Jinwu Hu et al.

Instead of picking training data based only on external metrics, you can use SAEs to decode what the model actually learns internally, then use those signals to organize data better—making training more efficient without changing the model architecture.

This paper shows how to improve LLM training by using Sparse Autoencoders (SAEs) to read the model's internal representations and guide data selection. The method clusters training data for diversity, orders it by difficulty, and filters low-quality examples—improving math performance by 3% and cutting training time by 20% on smaller models.

Zhifei Dou, Shabnam Hassani, Ou Wei

Adding simple edge detection to flowchart images helps VLMs understand topology better—a practical, training-free technique that improves industrial document processing by 11-17 percentage points without requiring annotated data.

EdgeFlow improves how Vision Language Models convert flowcharts into machine-readable formats by adding edge detection as a visual guide. The method works without training data or fine-tuning, achieving significant improvements on real-world industrial flowcharts by helping the model better understand the structure and connections between elements.

Mirac Suzgun, Emily Shen, Federico Bianchi et al.

AI chatbots excel at retrieving and synthesizing recent news but have three critical weaknesses: they systematically underperform on non-English content, fail primarily due to retrieval errors rather than reasoning mistakes, and are easily fooled by questions containing subtle false information.

This study evaluates six major AI chatbots (Gemini, Grok, Claude, GPT models) on their ability to answer factual news questions across six languages and regions.

Pilchen Hippolyte, Fabre Romain, Signe Talla Franck et al.

Training LLMs on chronologically ordered data instead of shuffled data improves their knowledge of recent facts and temporal accuracy, suggesting data ordering matters for building models that stay current.

This paper investigates how the order of training data affects what LLMs learn about time-sensitive facts. Researchers trained 6B-parameter models on chronologically ordered data versus shuffled data, and found that sequential training produces models with more current and accurate temporal knowledge while maintaining general language understanding.

Minrui Xu, Zilin Wang, Mengyi DENG et al.

Automated environment synthesis and trajectory generation can reduce the data requirements for tool-use agent training by 5x while improving downstream performance, making agentic RL more practical and scalable.

EnvFactory automates the creation of tool-use training environments and realistic multi-turn interaction trajectories for teaching language models to use tools effectively. It generates diverse, natural training data from verified executable environments, enabling more efficient agent training with fewer resources than existing approaches.

Yuwen Du, Rui Ye, Shuo Tang et al.

High-quality training data matters more than pipeline complexity: careful data curation with SFT alone can beat industrial-scale approaches combining pre-training, continual pre-training, and RL for building capable search agents.

OpenSeeker-v2 shows that simple supervised fine-tuning on carefully designed training data can match or beat complex industrial pipelines for building search agents.

Bogdan Oancea

Unsupervised learning can detect multivariate anomalies in regional data that traditional single-variable checks miss, helping statistical agencies distinguish between data quality issues and genuine structural divergence.

This paper uses five unsupervised machine learning techniques to detect regions in Europe with unusual combinations of economic and social indicators, rather than just extreme individual values.

Tanush Yadav, Mohammadreza Salehi, Jae Sung Park et al.

Vision-language models perform surprisingly poorly on domain-specific action recognition even in simplified settings, but fine-tuning on domain-specific video data significantly closes the gap.

VideoNet is a new benchmark and dataset for testing how well AI models recognize specific actions in videos across 37 different domains. The researchers found that current vision-language models struggle with domain-specific action recognition—even simple binary choices—and created a 500k video question-answer dataset to improve performance through fine-tuning.

Tao Ge, Baolin Peng, Hao Cheng et al.

Synthetic computer environments with long-horizon simulations can generate realistic training data for productivity agents at scale, enabling them to learn from diverse workplace scenarios without human annotation.

Researchers created a system to generate realistic computer environments at scale—complete with folder structures and documents—then simulated AI agents working on month-long productivity tasks within them.

Yujun Wu, Dongxu Zhang, Xinchen Li et al.

Structured knowledge of method evolution, not just citations, is essential infrastructure for AI agents doing research. This graph enables machines to understand how innovations emerge and build upon each other, unlocking automated idea evaluation and generation.

Intern-Atlas is a structured database of how AI research methods evolve and build on each other, extracted from over 1 million papers. Unlike traditional citation networks, it explicitly maps methodological relationships—showing which techniques led to which innovations and why—making it queryable for AI research agents and enabling automated discovery of new research directions.

Binghao Huang, Yunzhu Li

Practical tactile sensing for robotics is now accessible to researchers and developers without expensive custom hardware—FlexiTac provides a plug-and-play solution that integrates with standard robot learning pipelines.

FlexiTac is an affordable, open-source tactile sensor system for robot hands and grippers that combines flexible sensor pads with simple electronics to provide real-time touch feedback. It works with existing robot platforms and supports modern AI training methods like learning from combined vision and touch data.

An-Yang Ji, Jun-Peng Jiang, De-Chuan Zhan et al.

LLMs fail at implicit prediction tasks on tables because they don't recognize when a question requires inference from patterns rather than lookup; intent disambiguation is the critical bottleneck.

TopBench is a benchmark for testing how well language models can answer questions about tables that require prediction and reasoning, not just data lookup. It includes 779 examples across tasks like forecasting values, analyzing treatment effects, and complex filtering—revealing that current models struggle to recognize when prediction is needed and often default to simple retrieval instead.

Ansar Aynetdinov, Patrick Haller, Alan Akbik

For non-English language models, aggressively filtering data for quality and repeating it multiple times beats training once on larger, diverse datasets—a practical insight for resource-constrained language model development.

This paper challenges the assumption that diverse data is always better for language model training. For German, the researchers found that repeatedly training on a smaller, high-quality filtered dataset outperforms training once on a larger, less-filtered dataset—even after 7 epochs of repetition.

Nadine El-Naggar, Tatsuki Kuribayashi, Ted Briscoe

Curriculum learning substantially changes language models' learning biases, suggesting that training order matters as much as model architecture when predicting which language structures are 'easy' to learn.

This paper investigates how curriculum learning—training language models on simpler sentences first rather than random order—affects which linguistic patterns models naturally learn.

Griffin Pitts, Muntasir Hoq, Peter Brusilovsky et al.

By extracting knowledge components from student code patterns, you can steer generative models to create personalized learning content that directly targets the logical errors students are making, rather than relying on generic pre-written examples.

This paper presents a system that automatically generates personalized worked examples for programming students based on their actual code submissions. Instead of using fixed example libraries, the system analyzes patterns in student errors using code structure analysis and uses these patterns to guide an AI model to create relevant examples that address each student's specific misconceptions.

Nirmit Joshi, Roey Magen, Nathan Srebro et al.

Learning from diverse reasoning traces is harder than learning from a single thinker, but you can overcome this by actively collecting reasoning data from many thinkers (logarithmic in target accuracy) combined with passive final-answer supervision.

This paper studies how AI models can learn from multiple people or programs solving the same problem in different ways (e.g., different math solutions or code implementations).

Amal Akli, Mike Papadakis, Maxime Cordy et al.

Task description quality matters more than model size for reliable code generation—a small, fine-tuned classifier can detect problematic descriptions better than much larger models, and under-specification is the most critical defect type to watch for.

This paper introduces SpecValidator, a lightweight classifier that detects defects in task descriptions given to code-generating AI models. The tool identifies three types of problems—vague language, missing details, and formatting issues—and shows it's much better at catching these issues than larger models like GPT-4 mini or Claude.

Parthasarathi Panda, Asheswari Swain, Subhrakanta Panda

You can select optimal training data 40x faster than competing methods by matching source distributions through clustering and target distributions through regression, without sacrificing quality.

CRAFT is a fast method for selecting high-quality training data subsets from massive datasets. It uses clustering and statistical matching to pick training examples whose target outputs align with your validation set, enabling efficient fine-tuning of translation models on millions of examples in under a minute.

Flávio Soriano, Victoria F. Mello, Pedro B. Rigueira et al.

Political discourse analysis using NLP can reveal patterns invisible in voting records alone—like how stylistic shifts, topic priorities, and speaker alignments reflect deeper political structures beyond formal party lines.

This paper analyzes 450,000+ speeches from Brazil's Chamber of Deputies using computational methods to understand how politicians speak, what they discuss, and who aligns with whom rhetorically.

Sherly Alfonso-Sánchez, Cristián Bravo, Kristina G. Stankova

Geographic representation matters more than model complexity for insurance risk prediction—simple coordinate + environmental feature combinations often outperform complex image-based approaches in zone-level claim frequency models.

This study shows how to improve motor insurance claim prediction by adding geographic data to standard actuarial models, even when location information is limited. Researchers tested environmental features from maps and satellite imagery on insurance claims data, finding that combining coordinates with environmental data works best, while image embeddings help when map data isn't available.

Praval Sharma, Ashok Samal, Leen-Kiat Soh et al.

This dataset enables building event extraction systems that work across diverse real-world documents and geographical contexts, moving beyond closed-domain limitations that plagued previous approaches.

EVENT5Ws is a large, manually annotated dataset for extracting key event information (who, what, when, where, why) from documents in open-domain settings.

Yuto Nishida, Naoki Shikoda, Yosuke Kishinami et al.

LLMs don't memorize facts in a surface-invariant way; their ability to answer factual questions depends heavily on which name or spelling variant you use for an entity, suggesting memorization is tied to specific linguistic forms encountered during training.

This paper investigates how large language models memorize facts by testing whether they can answer questions about the same entity using different names and spellings.

Safouane El Ghazouali, Nicola Venturi, Michael Rueegsegger et al.

Synthetic data with perfect annotations can accelerate progress on multiple aerial imagery tasks simultaneously—depth, domain shift, and resolution—without the cost and difficulty of collecting real-world ground truth.

SyMTRS is a large synthetic dataset for aerial imagery that provides pixel-perfect depth maps, day/night image pairs, and multi-scale variants for training AI models on three tasks: depth estimation, domain adaptation, and super-resolution.

Yvon K. Awuklu, Meghyn Bienvenu, Katsumi Inoue et al.

You can build practical event detection systems using logical rules and constraint satisfaction that work efficiently on real timestamped data while handling conflicting inferences—demonstrated on medical records.

This paper presents a logic-based system for detecting high-level events from timestamped data, like inferring disease episodes from patient medical records. The system uses logical rules to identify events, handles conflicts between inferred events, and can run efficiently on real data while staying aligned with expert knowledge.

Hans Ole Hatzel, Ekaterina Artemova, Haimo Paul Stiemer et al.

Narrative similarity can be operationalized as a practical classification task, and LLM ensembles currently outperform other approaches, but there's significant room for improvement in how systems represent and compare story meanings.

This paper introduces a shared task on narrative similarity that asks systems to determine which of two stories is more similar to a reference story. The team collected over 1,000 annotated story triples and evaluated 71 submissions from 46 teams, finding that LLM ensembles performed best for classification while fine-tuned embedding models competed well with simpler approaches.

Breno Matos, Rennan C. Lima, Savvas Zannettou et al.

Misinformation detection is more useful when you know *where* in a video the false claim occurs, not just *whether* it exists—this work enables fine-grained detection at the segment level rather than video level.

This paper tackles video misinformation by identifying exactly where false claims appear within videos. Instead of just labeling entire videos as true or false, researchers transcribed video audio and annotated which specific segments contain misinformation, creating two datasets with 500+ videos. They trained language models to pinpoint these problematic spans, achieving 68% F1 score.

Tasnim Kabir, Dmytro Kurdydyk, Aadi Palnitkar et al.

Current audio AI models fail dramatically on genuine audio understanding tasks—they likely exploit dataset biases and metadata rather than actually listening to and reasoning about sound.

AUDITA is a new benchmark dataset with real-world audio and human-authored trivia questions designed to test whether AI models can truly understand audio content rather than relying on shortcuts. Humans answer correctly 32% of the time, but state-of-the-art models score below 9%, revealing a significant gap in audio reasoning capabilities.