MI-CXR is a new benchmark that shows state-of-the-art vision-language models achieve only 29.3% accuracy on longitudinal reasoning tasks across multi-visit chest X-ray sequences.
Title resolution pending
25 Pith papers cite this work. Polarity classification is still indexing.
25
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
DRIFTLENS quantifies memory-induced reasoning drift in personalized LLMs, finding medium-to-large effects across four models and ten user attributes that post-training only partly reduces.
OLIVIA treats LLM agent action selection as a contextual linear bandit over frozen hidden states and applies UCB exploration to adapt online, yielding consistent gains over static ReAct and prompt-based baselines on four benchmarks.
Fully Looped Transformer stabilizes looped training up to 12 iterations via distributed inter-loop signals and attention injection, improving downstream performance by up to 13.2%.
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
Code LLMs generate substantially worse comments outside English, and no tested automatic metric or LLM judge reliably matches human assessment of those outputs.
Prepending stochastic sequences from Lorem Ipsum vocabulary to prompts during GRPO resampling broadens reasoning exploration and outperforms standard resampling on hard tasks for 1.7B-7B models.
Omni-MATH supplies 4428 human-verified Olympiad math problems that expose top LLMs achieving only 52.55% to 60.54% accuracy on the most difficult items.
MultiHop-RAG is a new benchmark dataset demonstrating that existing retrieval-augmented generation systems perform poorly on multi-hop queries requiring retrieval and reasoning over multiple evidence pieces.
Attractor Models solve for fixed points in transformer embeddings using implicit differentiation to enable stable iterative refinement, delivering better perplexity, accuracy, and efficiency than standard or looped transformers.
Structured Recurrent Mixers provide a dual parallel-recurrent representation for sequence models, claiming superior training efficiency, information capacity, and inference throughput over linear complexity alternatives.
BioTool dataset enables fine-tuning a 4B-parameter LLM to outperform GPT-5.1 in biomedical tool calling while improving downstream answer quality per human experts.
Spherical flows on S^{d-1} with vMF noise reduce the continuity equation to a scalar ODE in cosine similarity, yielding posterior-weighted marginal velocity and score that enable ODE and predictor-corrector sampling for categorical sequences, with the posterior trained by cross-entropy and empirical
Fine-tuning LLMs on Arabic yields similar zero-shot gains on Semitic and non-Semitic languages, with chain-of-thought reasoning producing parallel benefits, indicating task alignment drives transfer more than language relatedness.
DiffMAS jointly optimizes latent communication and reasoning in multi-agent LLM systems via parameter-efficient supervised training on trajectories, yielding consistent gains over baselines on math, science, and code benchmarks.
CAL-GRPO calibrates per-attempt weights in multi-attempt CoT to deliver unbiased gradients for optimizing Verification@K success while keeping variance low.
DPOP is a new loss function that prevents DPO from lowering preferred response likelihoods and outperforms standard DPO on diverse datasets, MT-Bench, and enables Smaug-72B to exceed 80% on the Open LLM Leaderboard.
Continued pretraining of Code Llama on Proof-Pile-2 yields Llemma, an open math-specialized LLM that beats known open base models on MATH and supports tool use plus formal proving out of the box.
Chain-of-Verification reduces hallucinations in large language models by drafting responses, planning independent verification questions, answering them separately, and generating a final verified output.
FaithMed applies reinforcement learning with process-level rewards derived from evidence-based medicine rubrics to improve both task performance and reasoning faithfulness in medical LLMs.
R2IF uses a composite reward (binary correctness, CoT effectiveness, and parameter-level SMV) under GRPO to align LLM reasoning with function-calling decisions, improving accuracy and reasoning quality on BFCL/ACEBench.
JTPRO co-optimizes prompts and tool descriptions via reflection to raise overall success rate by 5-20% over baselines on multi-tool benchmarks.
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
Hy-MT2 presents three new multilingual translation models that claim to outperform listed open-source and commercial systems on diverse tasks while enabling low-storage on-device use.
citing papers explorer
-
MI-CXR: A Benchmark for Longitudinal Reasoning over Multi-Interval Chest X-rays
MI-CXR is a new benchmark that shows state-of-the-art vision-language models achieve only 29.3% accuracy on longitudinal reasoning tasks across multi-visit chest X-ray sequences.
-
DRIFTLENS: Measuring Memory-Induced Reasoning Drift in Personalized Language Models
DRIFTLENS quantifies memory-induced reasoning drift in personalized LLMs, finding medium-to-large effects across four models and ten user attributes that post-training only partly reduces.
-
OLIVIA: Online Learning via Inference-time Action Adaptation for Decision Making in LLM ReAct Agents
OLIVIA treats LLM agent action selection as a contextual linear bandit over frozen hidden states and applies UCB exploration to adapt online, yielding consistent gains over static ReAct and prompt-based baselines on four benchmarks.
-
Simply Stabilizing the Loop via Fully Looped Transformer
Fully Looped Transformer stabilizes looped training up to 12 iterations via distributed inter-loop signals and attention injection, improving downstream performance by up to 13.2%.
-
Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
-
Evaluating Non-English Developer Support in Machine Learning for Software Engineering
Code LLMs generate substantially worse comments outside English, and no tested automatic metric or LLM judge reliably matches human assessment of those outputs.
-
Nonsense Helps: Prompt Space Perturbation Broadens Reasoning Exploration
Prepending stochastic sequences from Lorem Ipsum vocabulary to prompts during GRPO resampling broadens reasoning exploration and outperforms standard resampling on hard tasks for 1.7B-7B models.
-
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
Omni-MATH supplies 4428 human-verified Olympiad math problems that expose top LLMs achieving only 52.55% to 60.54% accuracy on the most difficult items.
-
MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries
MultiHop-RAG is a new benchmark dataset demonstrating that existing retrieval-augmented generation systems perform poorly on multi-hop queries requiring retrieval and reasoning over multiple evidence pieces.
-
Solve the Loop: Attractor Models for Language and Reasoning
Attractor Models solve for fixed points in transformer embeddings using implicit differentiation to enable stable iterative refinement, delivering better perplexity, accuracy, and efficiency than standard or looped transformers.
-
Structured Recurrent Mixers for Massively Parallelized Sequence Generation
Structured Recurrent Mixers provide a dual parallel-recurrent representation for sequence models, claiming superior training efficiency, information capacity, and inference throughput over linear complexity alternatives.
-
BioTool: A Comprehensive Tool-Calling Dataset for Enhancing Biomedical Capabilities of Large Language Models
BioTool dataset enables fine-tuning a 4B-parameter LLM to outperform GPT-5.1 in biomedical tool calling while improving downstream answer quality per human experts.
-
Spherical Flows for Sampling Categorical Data
Spherical flows on S^{d-1} with vMF noise reduce the continuity equation to a scalar ODE in cosine similarity, yielding posterior-weighted marginal velocity and score that enable ODE and predictor-corrector sampling for categorical sequences, with the posterior trained by cross-entropy and empirical
-
Disentangling Linguistic Relatedness from Task Alignment in Cross-Lingual Transfer
Fine-tuning LLMs on Arabic yields similar zero-shot gains on Semitic and non-Semitic languages, with chain-of-thought reasoning producing parallel benefits, indicating task alignment drives transfer more than language relatedness.
-
Learning to Communicate: Toward End-to-End Optimization of Multi-Agent Language Systems
DiffMAS jointly optimizes latent communication and reasoning in multi-agent LLM systems via parameter-efficient supervised training on trajectories, yielding consistent gains over baselines on math, science, and code benchmarks.
-
Learning to Correct: Calibrated Reinforcement Learning for Multi-Attempt Chain-of-Thought
CAL-GRPO calibrates per-attempt weights in multi-attempt CoT to deliver unbiased gradients for optimizing Verification@K success while keeping variance low.
-
Smaug: Fixing Failure Modes of Preference Optimisation with DPO-Positive
DPOP is a new loss function that prevents DPO from lowering preferred response likelihoods and outperforms standard DPO on diverse datasets, MT-Bench, and enables Smaug-72B to exceed 80% on the Open LLM Leaderboard.
-
Llemma: An Open Language Model For Mathematics
Continued pretraining of Code Llama on Proof-Pile-2 yields Llemma, an open math-specialized LLM that beats known open base models on MATH and supports tool use plus formal proving out of the box.
-
Chain-of-Verification Reduces Hallucination in Large Language Models
Chain-of-Verification reduces hallucinations in large language models by drafting responses, planning independent verification questions, answering them separately, and generating a final verified output.
-
FaithMed: Training LLMs For Faithful Evidence-Based Medical Reasoning
FaithMed applies reinforcement learning with process-level rewards derived from evidence-based medicine rubrics to improve both task performance and reasoning faithfulness in medical LLMs.
-
R2IF: Aligning Reasoning with Decisions via Composite Rewards for Interpretable LLM Function Calling
R2IF uses a composite reward (binary correctness, CoT effectiveness, and parameter-level SMV) under GRPO to align LLM reasoning with function-calling decisions, improving accuracy and reasoning quality on BFCL/ACEBench.
-
JTPRO: A Joint Tool-Prompt Reflective Optimization Framework for Language Agents
JTPRO co-optimizes prompts and tool descriptions via reflection to raise overall success rate by 5-20% over baselines on multi-tool benchmarks.
-
Bad Seeing or Bad Thinking? Rewarding Perception for Multimodal Reasoning
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
-
Hy-MT2: A Family of Fast, Efficient and Powerful Multilingual Translation Models in the Wild
Hy-MT2 presents three new multilingual translation models that claim to outperform listed open-source and commercial systems on diverse tasks while enabling low-storage on-device use.
- Lessons from the Trenches on Reproducible Evaluation of Language Models