Implement a reasoning LLM in PyTorch from scratch, step by step

(ArXiv25) Vision Matters: Simple Visual Perturbations Can Boost Multimodal Math Reasoning

Solving Inequality Proofs with Large Language Models.

An official implementation of "SPARK: Synergistic Policy And Reward Co-Evolving Framework"

Prompt-engineering study on LLM math reasoning (GSM8K) and code generation (HumanEval): zero/few-shot, self-consistency, self-verification, and experiments on prompt quality, complexity, demonstrations, and diversity.

ArmLLM 2025 solutions covering ViT from scratch, SigLIP–Qwen LaTeX OCR, GRPO reasoning post-training, inference-time reasoning strategies, and adversarial vision attacks.

AI Benchmark 知识库 — 全面收录各大 AI 公司用来测试模型性能的 Benchmark 题库完整集合

A minimal JEPA-based language model demonstrating latent-space reasoning on GSM8K using a single decoder-only Transformer.

STaR × S1 math pipeline on Qwen2.5-1.5B. LoRA, strict Final: format, ~20–30% acc (OpenR1-Math split).

Data cleaning and structuring pipeline for math reasoning tasks using Qwen3-0.6B for LLM post-training.

A controlled LoRA finetuning study on process supervision for mathematical reasoning with Qwen2.5-Math-7B-Instruct.

Paired-framing pilot on observable self-doubt in math reasoning models

Beyond English-Only GRPO: a multi-seed controlled study of training-language and auxiliary-reward effects in sub-3B math reasoning (GRPO + LoRA, single GPU).

GRPO (Group Relative Policy Optimization) implemented from scratch in PyTorch. 10 ablation experiments.

Verifier-backed math reasoning lab for proof flaw localization, minimal repair, and RL-style evaluation.

NLP course final project (2026), Nanjing Normal University, supervised by 孔力: GSM8K math QA with Seq2Seq, Transformer and LLMs.

MSc capstone: mechanistic interpretability of Chain-of-Thought reasoning in LLMs via SAE features and logprob signals

Comprehensive framework for mathematical reasoning research with dual research capabilities

Small-scale Implementation and Extension of “The Surprising Effectiveness of Negative Reinforcement in LLM Reasoning” (NeurIPS '25)

Unofficial PyTorch reproduction for LIMO: Less is More for Reasoning.