Stars
Beyond Majority Voting: Towards Fine-grained and More Reliable Reward Signal for Test-Time Reinforcement
SemPA: Improving Sentence Embeddings of Large Language Models through Semantic Preference Alignment
Official implementation of the NeurIPS 2025 paper "Soft Thinking: Unlocking the Reasoning Potential of LLMs in Continuous Concept Space"
This repository contains a regularly updated paper list for LLMs-reasoning-in-latent-space.
[TMLR 2026] Survey: https://arxiv.org/pdf/2507.20198
Enhancing Automated Interpretability with Output-Centric Feature Descriptions
Attribute-guided reinforcement learning framework for molecular property prediction with large language models.
Code for the ICML 2025 Paper "Product of Experts with LLMs: Boosting Performance on ARC is a Matter of Perspective"
This repository contains the official implementation of the paper **"Improving Rationality in the Reasoning Process of Language Models through Self-playing Game."**
TextGrad: Automatic ''Differentiation'' via Text -- using large language models to backpropagate textual gradients. Published in Nature.
[ACL'25 Findings] LDIR (Low-Dimensional Dense Interpretable Text Embeddings with Relative Representations) is a novel text embedding method that balances semantic expressiveness, interpretability, …
[ACL'25 Findings] RankedVotingSC (Ranked Voting based Self-Consistency) is a method that generates ranked answers in each reasoning attempt and aggregates them using ranked voting across multiple r…
A Survey of Reinforcement Learning for Large Reasoning Models
国家自然科学基金申请书正文（面上项目）LaTeX 模板（非官方）
Our solution for the arc challenge 2024
Public repository for "The Surprising Effectiveness of Test-Time Training for Abstract Reasoning"
Implementation for ACL 2024 paper "Meta-Task Prompting Elicits Embeddings from Large Language Models"
[ECCV 2024 Best Paper Candidate & TPAMI 2025] PointLLM: Empowering Large Language Models to Understand Point Clouds
LAVIS - A One-stop Library for Language-Vision Intelligence
Code for "Visual Spatial Description: Controlled Spatial-Oriented Image-to-Text Generation"
[ICCV 2025] Improving 3D Large Language Model via Robust Instruction Tuning