MedPRMBench is the first fine-grained benchmark for process reward models in medical reasoning, featuring 6500 questions, 13000 chains, 113910 step labels, and a baseline that improves downstream QA accuracy by 3.2-6.7 points.
hub
Marco-o1: Towards open reasoning models for open-ended solutions
16 Pith papers cite this work. Polarity classification is still indexing.
16
Pith papers citing it
hub tools
citation-role summary
background 3
baseline 1
citation-polarity summary
representative citing papers
Behavior Forecasters trained on LRM trajectories outperform larger models in predicting repeatability and input sensitivity at low cost.
TVI-CoT introduces learnable control tokens <THINK>, <LOOK>, <ANSWER> that let multimodal LLMs interleave textual reasoning with dynamic visual feature access, reporting gains of 3.4-6.1% on eight benchmarks over prior CoT baselines.
Document-level machine translation followed by segment-level LLM refinement provides the strongest and most stable improvements in literary translation quality, mainly enhancing fluency and style rather than adequacy.
ReflectMT internalizes reflection via two-stage RL to enable direct high-quality machine translation that outperforms explicit reasoning models like DeepSeek-R1 on WMT24 while using 94% fewer tokens.
MADE is a new multilingual agentic diagnosing engine that produces higher-quality diagnostic reports (47% better than baseline) on a large-scale evaluation substrate covering 33 model families and 26 languages.
Survey that defines agentic RL for LLMs via POMDPs, introduces a taxonomy of planning/tool-use/memory/reasoning capabilities and domains, and compiles open environments from over 500 papers.
ZeroSearch uses supervised fine-tuning to create a simulated retrieval module and curriculum-based RL rollouts that degrade document quality to train LLMs on search capabilities without real search API calls.
R1-Searcher uses two-stage outcome-based RL to train LLMs to invoke external search systems for better reasoning without process rewards or distillation.
HuatuoGPT-o1 achieves superior medical complex reasoning by using a verifier to curate reasoning trajectories for fine-tuning and then applying RL with verifier-based rewards.
A survey of test-time scaling for multimodal foundation models that introduces a three-way taxonomy of sampling, feedback, and search approaches along with applications and benchmarks.
MARS-SQL trains a multi-agent RL system with ReAct-style interaction and generative validation to produce SQL queries, reaching 77.84% execution accuracy on BIRD dev and 89.75% on Spider test.
MODE-RAG introduces a VFE-driven multi-agent pipeline with MCTS and logit perturbations to lower hallucination and sycophancy rates in multimodal RAG, tested on the new ModeVent subset of MultiVent.
The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.
STILL-2 uses imitation of distilled long-form thoughts, multi-rollout exploration on difficult problems, and iterative self-improvement of the dataset to train reasoning models that reach competitive performance on three challenging benchmarks.
The paper provides the first comprehensive survey of multimodal chain-of-thought reasoning, including foundational concepts, a taxonomy of methodologies, application analyses, challenges, and future directions.
citing papers explorer
-
MedPRMBench: A Fine-grained Benchmark for Process Reward Models in Medical Reasoning
MedPRMBench is the first fine-grained benchmark for process reward models in medical reasoning, featuring 6500 questions, 13000 chains, 113910 step labels, and a baseline that improves downstream QA accuracy by 3.2-6.7 points.
-
Forecasting Future Behavior as a Learning Task
Behavior Forecasters trained on LRM trajectories outperform larger models in predicting repeatability and input sensitivity at low cost.
-
TVI-CoT: Text-Visual Interleaved Chain-of-Thought Reasoning for Multimodal Understanding
TVI-CoT introduces learnable control tokens <THINK>, <LOOK>, <ANSWER> that let multimodal LLMs interleave textual reasoning with dynamic visual feature access, reporting gains of 3.4-6.1% on eight benchmarks over prior CoT baselines.
-
What Does LLM Refinement Actually Improve? A Systematic Study on Document-Level Literary Translation
Document-level machine translation followed by segment-level LLM refinement provides the strongest and most stable improvements in literary translation quality, mainly enhancing fluency and style rather than adequacy.
-
ReflectMT: Internalizing Reflection for Efficient and High-Quality Machine Translation
ReflectMT internalizes reflection via two-stage RL to enable direct high-quality machine translation that outperforms explicit reasoning models like DeepSeek-R1 on WMT24 while using 94% fewer tokens.
-
MADE: Beyond Scoring via a Multilingual Agentic Diagnosing Engine for Fine-Grained Evaluation Insights
MADE is a new multilingual agentic diagnosing engine that produces higher-quality diagnostic reports (47% better than baseline) on a large-scale evaluation substrate covering 33 model families and 26 languages.
-
The Landscape of Agentic Reinforcement Learning for LLMs: A Survey
Survey that defines agentic RL for LLMs via POMDPs, introduces a taxonomy of planning/tool-use/memory/reasoning capabilities and domains, and compiles open environments from over 500 papers.
-
ZeroSearch: Incentivize the Search Capability of LLMs without Searching
ZeroSearch uses supervised fine-tuning to create a simulated retrieval module and curriculum-based RL rollouts that degrade document quality to train LLMs on search capabilities without real search API calls.
-
R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning
R1-Searcher uses two-stage outcome-based RL to train LLMs to invoke external search systems for better reasoning without process rewards or distillation.
-
HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs
HuatuoGPT-o1 achieves superior medical complex reasoning by using a verifier to curate reasoning trajectories for fine-tuning and then applying RL with verifier-based rewards.
-
Test-Time Scaling in Multimodal Foundation Models: A Comprehensive Survey of Generation and Reasoning
A survey of test-time scaling for multimodal foundation models that introduces a three-way taxonomy of sampling, feedback, and search approaches along with applications and benchmarks.
-
MARS-SQL: A multi-agent reinforcement learning framework for Text-to-SQL
MARS-SQL trains a multi-agent RL system with ReAct-style interaction and generative validation to produce SQL queries, reaching 77.84% execution accuracy on BIRD dev and 89.75% on Spider test.
-
MODE-RAG: Manifold Outlier Diagnosis and Energy-based Retrieval-Augmented Generation Evaluation
MODE-RAG introduces a VFE-driven multi-agent pipeline with MCTS and logit perturbations to lower hallucination and sycophancy rates in multimodal RAG, tested on the new ModeVent subset of MultiVent.
-
From System 1 to System 2: A Survey of Reasoning Large Language Models
The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.
-
Imitate, Explore, and Self-Improve: A Reproduction Report on Slow-thinking Reasoning Systems
STILL-2 uses imitation of distilled long-form thoughts, multi-rollout exploration on difficult problems, and iterative self-improvement of the dataset to train reasoning models that reach competitive performance on three challenging benchmarks.
-
Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey
The paper provides the first comprehensive survey of multimodal chain-of-thought reasoning, including foundational concepts, a taxonomy of methodologies, application analyses, challenges, and future directions.