1.Ikaleio / lm-detector
Model fingerprint detection with a shared Web and Bun CLI reference bank
+501 stars
from 1,132 · score 6.22
Trending repositories
Ranked by Attention Score: stars gained over the last 7 complete days
17 repositories by Attention Score · Oct 29, 2026 – Nov 4, 2026 (UTC)
2.lyramakesmusic / neuron-explorer
Steering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout, browse the results
- interpretability
- mechanistic-interpretability
- qwen
- steering
+301 stars
from 1,613 · score 5.71
3.promptfoo / promptfoo
Test your prompts, agents, and RAGs. Red teaming/pentesting/vulnerability scanning for AI. Compare performance of GPT, Claude, Gemini, DeepSeek, and more. Simple declarative configs with command line and CI/CD integration. Used by OpenAI and Anthropic.
- llm
- prompt-engineering
- prompts
- llmops
- prompt-testing
- testing
- +12 more
+281 stars
from 26,562 · score 5.64
+256 stars
from 5 · score 5.55
5.rsi-gear / gear
Scikit-learn for Recursive Self-improvement (RSI). A modular framework for recursive harness evolution.
- meta-learning
- reinforcement-learning
- rsi
- ai-agents
- continual-learning
- infrastructure
- +6 more
+119 stars
from 295 · score 4.79
6.Shyboy0499 / precondition-library
An agent that compiles git maintenance chores into reusable programs, then decides whether to replay them by running checks rather than by guessing the task looks familiar
- agent-architecture
- git-automation
- llm-agents
- program-synthesis
- research
+109 stars
from 233 · score 4.70
7.slee-persis / GVS5H
GVS5H: Five Qwen3.8-27B Models Match Claude Fable 5 on LiveCodeBench Hard | Fable 5 Level Coding for a Fifth the Price - or on a Single GPU
+97 stars
from 1,095 · score 4.58
8.MCG-NJU / AutoTavern
TavernGym and TavernBench: a faithful, full-content Hearthstone Battlegrounds simulator with an RL interface, and a 960-task decision benchmark. Code and data for "Build, Don't Play".
- game-ai
- hearthstone
- reinforcement-learning
- rust
+91 stars
from 145 · score 4.52
9.Ta0p1 / ResolveLab
Fault-injected tool environments with side-effect grading for evaluating and post-training LLM agents
+64 stars
from 42 · score 4.17
10.EleutherAI / lm-evaluation-harness
A framework for few-shot evaluation of language models.
- evaluation-framework
- language-model
- transformer
+59 stars
from 14,342 · score 4.09
11.mll-lab-nu / embodied-first-exam
The first exam for code agents in the physical world: frozen robot tasks from many benchmark suites, each answered with a program and verified in a separate sandbox. Runs on Harbor.
- ai-agents
- benchmark
- coding-agents
- embodied-ai
- evaluation
- harbor
- +4 more
+29 stars
from 205 · score 3.40
+23 stars
from 426 · score 3.18
13.superwesleyhys-ux / factcircuit
FactCircuit: an auditable claim-and-evidence verification loop with immutable provenance, staged checks, bounded retrieval, and reproducible evaluation.
- accuracy-decline
- ai-safety
- audit-trail
- fact-checking
- information-retrieval
- llm-evaluation
- +5 more
+21 stars
from 905 · score 3.09
+21 stars
from 291 · score 3.09
15.latitude-dev / latitude-llm
Open-source observability for AI agents. Find where your agents fail, dispatch your coding agent to fix it, and verify the fix against real traces.
- llm-observability
- agent-monitoring
- agent-observability
- ai-error-monitor
- ai-monitoring
- ai-observability
- +2 more
+6 stars
from 4,739 · score 1.95
16.microsoft / promptflow
Build high-quality LLM apps - from prototyping, testing to production deployment and monitoring.
- ai
- llm
- chatgpt
- gpt
- prompt
- prompt-engineering
- +2 more
+4 stars
from 11,256 · score 1.61
17.paper-instruments / feather-train
Feather agent harness, training integrations, and evaluation tools for knowledge work
+2 stars
from 226 · score 1.10