Official Repository: A Comprehensive Benchmark for Logical Reasoning in MLLMs
-
Updated
Jun 17, 2025 - Python
#
reasoning-benchmark
Here are 8 public repositories matching this topic...
Live Deep Research Bench. A challenging, objective benchmark for deep research tasks.
-
Updated
Oct 16, 2025 - Python
First-Person Agent Memory Bench. 10 Categories including fact recall, multi-hop links, temporal reasoning, fact overwrites, speaker traps, refusal, credibility, and agentic tool usage. 540K token / 60 session corpus, all in first person. Dynamic output-answer-key portion. Comprehensive report with visuals and miss breakdown.
ai-safety memory-benchmark ai-agent ai-governance ai-system-design reasoning-benchmark longmemeval ai-agent-memory ai-memory-system ai-agent-testing tool-use-behavior-identification beam-benchmark ai-agent-benchmark locomo-benchmark ai-memory-management context-memory-benchmark
-
Updated
Aug 27, 2026 - Python
Benchmark for evaluating advanced reasoning, recursive dependency resolution, and robustness capabilities of large language models in dynamic, noisy, and structurally challenging environments.
benchmark dependency-resolution ai-accuracy multistep-reasoning ai-evaluation large-language-models ai-reasoning llm-evaluation reasoning-benchmark llm-benchmark ai-reliability recursive-reasoning ai-stability long-chain-reasoning
-
Updated
May 15, 2026 - Python
Non-Western symbolic reasoning benchmark for LLMs. Multi-step rule-following inference within the Ba Zi (八字) formal system. Frozen lookup tables, Python reference implementation, mechanically verified gold CoT cases.
python nlp ai evaluation chinese-nlp symbolic-reasoning ba-zi rule-based-system multi-step-reasoning llm chain-of-thought llm-evaluation reasoning-benchmark llm-benchmark
-
Updated
May 13, 2026 - Python
Local-first benchmark for measuring reasoning improvement in small language models through critique and debate.
python slm huggingface llama-cpp ollama llm-evaluation self-critique small-language-models langgraph multi-agent-debate reasoning-benchmark adversarial-reasoning
-
Updated
Jun 20, 2026 - Python
Public preview of Society of Thought, a Qwen adapter that reasons through visible multi-persona debate, with benchmark evidence, raw traces, and demo.
benchmark evaluation transformer debate language-model gradio reasoning llm qwen reasoning-benchmark multi-persona reasoning-trace
-
Updated
Apr 1, 2026 - Python
CandyBench Notebook: visually select OpenAI-compatible relay models, run 10-way reasoning tests, and preserve every raw reply.
cli benchmark ai jupyter ipywidgets model-evaluation reasoning llm openai-compatible reasoning-benchmark veridrop
-
Updated
Aug 16, 2026 - Jupyter Notebook
Add this topic to your repo
To associate your repository with the reasoning-benchmark topic, visit your repo's landing page and select "manage topics."