Papers131

#reinforcement learning

Self-Improving Multilingual Long Reasoning via Translation-Reasoning Integrated Training

Junxiao Liu, Zhijun Wang et al.Feb 5arXiv

TRIT is a new training method that teaches AI to translate and think at the same time so it can solve hard problems in many languages without extra helper models.

#multilingual reasoning#translation-reasoning integration#self-translation

Skin Tokens: A Learned Compact Representation for Unified Autoregressive Rigging

Intermediate

Jia-peng Zhang, Cheng-Feng Pu et al.Feb 4arXiv

Rigging 3D characters is a bottleneck: making bones and skin weights by hand is slow and tricky, and past automatic tools often guess the skin weights poorly.

#auto-rigging#skinning weights#SkinTokens

ERNIE 5.0 Technical Report

Intermediate

Haifeng Wang, Hua Wu et al.Feb 4arXiv

ERNIE 5.0 is a single giant model that can read and create text, images, video, and audio by predicting the next pieces step by step, like writing a story one line at a time.

#ERNIE 5.0#unified autoregressive model#mixture-of-experts

Agent-Omit: Training Efficient LLM Agents for Adaptive Thought and Observation Omission via Agentic Reinforcement Learning

Intermediate

Yansong Ning, Jun Fang et al.Feb 4arXiv

Agent-Omit teaches AI agents to skip unneeded thinking and old observations, cutting tokens while keeping accuracy high.

#LLM agents#reinforcement learning#agentic RL

Reasoning Cache: Continual Improvement Over Long Horizons via Short-Horizon RL

Intermediate

Ian Wu, Yuxiao Qu et al.Feb 3arXiv

Reasoning Cache (RC) is a new way for AI to think in steps: it writes some thoughts, makes a short summary, throws away the long thoughts, and then keeps going using only the summary.

#Reasoning Cache#iterative decoding#summary-conditioned reasoning

Search-R2: Enhancing Search-Integrated Reasoning via Actor-Refiner Collaboration

Intermediate

Bowei He, Minda Hu et al.Feb 3arXiv

This paper teaches AI to look things up on the web and fix its own mistakes mid-thought instead of starting over from scratch.

#search-integrated reasoning#reinforcement learning#credit assignment

Learning Query-Specific Rubrics from Human Preferences for DeepResearch Report Generation

Intermediate

Changze Lv, Jie Zhou et al.Feb 3arXiv

DeepResearch agents write long, evidence-based reports, but teaching and grading them is hard because there is no single 'right answer' to score against.

#DeepResearch#query-specific rubrics#human preference learning

SWE-World: Building Software Engineering Agents in Docker-Free Environments

Intermediate

Shuang Sun, Huatong Song et al.Feb 3arXiv

SWE-World lets code-fixing AI agents practice and learn without heavy Docker containers by using smart models that pretend to be the computer and tests.

#SWE-World#software engineering agents#Docker-free training

SWE-Master: Unleashing the Potential of Software Engineering Agents via Post-Training

Intermediate

Huatong Song, Lisheng Huang et al.Feb 3arXiv

SWE-Master is a fully open, step-by-step recipe for turning a regular coding model into a strong software-fixing agent that works across many steps, files, and tests.

#SWE-Master#software engineering agent#long-horizon SFT

Self-Hinting Language Models Enhance Reinforcement Learning

Intermediate

Baohao Liao, Hanze Dong et al.Feb 3arXiv

When rewards are rare, a popular training method for language models (GRPO) often stops learning because every try in a group gets the same score, so there is nothing to compare.

#reinforcement learning#GRPO#self-hinting

Neural Predictor-Corrector: Solving Homotopy Problems with Reinforcement Learning

Intermediate

Jiayao Mai, Bangyan Liao et al.Feb 3arXiv

This paper shows that many hard math and AI problems can be solved with one shared idea called homotopy, where we move from an easy version of a problem to the real one step by step.

#homotopy continuation#predictor-corrector#reinforcement learning

Training LLMs for Divide-and-Conquer Reasoning Elevates Test-Time Scalability

Intermediate

Xiao Liang, Zhong-Zhi Li et al.Feb 2arXiv

The paper trains language models to solve hard problems by first breaking them into smaller parts and then solving those parts, instead of only thinking in one long chain.

#divide-and-conquer reasoning#chain-of-thought#reinforcement learning

1 2 3 4 5