(ArXiv25) Vision Matters: Simple Visual Perturbations Can Boost Multimodal Math Reasoning
-
Updated
Sep 30, 2025 - Python
#
math-reasoning
Here are 18 public repositories matching this topic...
Solving Inequality Proofs with Large Language Models.
-
Updated
Dec 15, 2025 - Python
An official implementation of "SPARK: Synergistic Policy And Reward Co-Evolving Framework"
self-improvement multi-modal large-language-models vision-language-model reward-model large-vision-language-models self-rewarding math-reasoning
-
Updated
Oct 23, 2025 - Python
ArmLLM 2025 solutions covering ViT from scratch, SigLIP–Qwen LaTeX OCR, GRPO reasoning post-training, inference-time reasoning strategies, and adversarial vision attacks.
reinforcement-learning computer-vision transformers pytorch multimodal-learning post-training adversarial-attacks adversarial-robustness vision-transformer latex-ocr vision-language-model qwen siglip inference-time-compute grpo math-reasoning armllm llm-summer-school
-
Updated
Nov 26, 2025 - Jupyter Notebook
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
benchmark knowledge-base model-evaluation reasoning multimodal ai-benchmarks instruction-following llm long-context safety-evaluation ai-performance math-reasoning coding-benchmark benchmark-collection eval-frameworks
-
Updated
Apr 16, 2026
A minimal JEPA-based language model demonstrating latent-space reasoning on GSM8K using a single decoder-only Transformer.
deep-learning pytorch transformer research-project representation-learning language-model latent-space gsm8k jepa math-reasoning
-
Updated
Feb 28, 2026 - Python
A controlled LoRA finetuning study on process supervision for mathematical reasoning with Qwen2.5-Math-7B-Instruct.
-
Updated
Apr 23, 2026 - Python
STaR × S1 math pipeline on Qwen2.5-1.5B. LoRA, strict Final: format, ~20–30% acc (OpenR1-Math split).
-
Updated
Sep 6, 2025 - Python
GSM8K benchmark results for Qwen3-4B and Qwen3-8B fine-tuned
-
Updated
Feb 28, 2026 - Python
Tool-Integrated Reasoning for competition math — weighted voting, difficulty-aware allocation
-
Updated
May 9, 2026 - Python
GRPO (Group Relative Policy Optimization) implemented from scratch in PyTorch. 10 ablation experiments.
-
Updated
Apr 26, 2026 - Python
Data cleaning and structuring pipeline for math reasoning tasks using Qwen3-0.6B for LLM post-training.
nlp machine-learning transformers pytorch data-processing data-cleaning post-training llm qwen math-reasoning
-
Updated
Apr 13, 2026 - Python
NDA-safe excerpts of math & economics modeling tasks for LLM reasoning evaluation and numerical verification.
python monte-carlo root-finding numerical-methods model-validation economic-modeling quantitative-research llm-evaluation math-reasoning
-
Updated
Dec 23, 2025 - Python
GRPO reinforcement learning with verifiable rewards for sub-2B models
-
Updated
May 9, 2026 - Python
Transforming weak prompts into reasoning machines using Textual Gradients and AdalFlow. Runs on Colab.
transformers pytorch google-colab large-language-models llm prompt-engineering chain-of-thought generative-ai gsm8k prompt-optimization textual-gradients automated-prompt-engineering math-reasoning llm-autodiff adalflow
-
Updated
Jan 28, 2026 - Python
Comprehensive framework for mathematical reasoning research with dual research capabilities
-
Updated
Mar 18, 2026 - Python
Small-scale Implementation and Extension of “The Surprising Effectiveness of Negative Reinforcement in LLM Reasoning” (NeurIPS '25)
-
Updated
Oct 23, 2025 - Python
Improve this page
Add a description, image, and links to the math-reasoning topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the math-reasoning topic, visit your repo's landing page and select "manage topics."