Контроль ризонинга и действий TUI-агентов по локальным трейсам: 8 скиллов (SKILL.md), сканер JSONL, контракт границ задачи · Trace-based control of terminal AI agents' reasoning and actions: 8 skills, JSONL scanner, task-boundary contracts
Self-healing code reasoning engine. Detective → QA → Patcher closed loop on SWE-bench, with an RL layer that turns every reasoning trace into DPO training data.
[CVPR 2026] StaR-KVQA: Structured Reasoning Traces for Implicit-Knowledge Visual Question Answering
🥈 2nd Place Winner | Adaption AutoScientist Challenge ($50,000 Prize Pool) — 4 specialized LLMs (70B, 109B, 120B) with LoRA fine-tuning, verified reasoning & synthetic datasets
Research codebase for an IEEE FLLM 2025 paper: retrieval-augmented LLM agent for CTF challenge solving with RAG, graph memory, tool routing, and reasoning traces.
Creating datasets and training code and dependancies.
Live LLM crypto trading competition data — daily decisions with full reasoning traces, trades, and equity curves for GPT, Claude, Gemini, Grok, DeepSeek and more. Benchmark data for comparing how LLMs trade. From traderank.ai
Visual Flamegraph & Cognitive Uncertainty Profiler for Agent Extended Thinking Traces (Claude Opus 5.5, GPT-6 Astra, Gemini 3.8 Flash)
Failure-aware audit protocol for reasoning-trajectory analysis in HumanEval code generation.
Zero-Knowledge Context Sanitizer & Token Blinded Encryption for Reasoning Traces (Claude Opus 5.5, GPT-6 Astra, Gemini 3.8 Flash)
To associate your repository with the reasoning-traces topic, visit your repo's landing page and select "manage topics."