llm-uncertainty

Here are 2 public repositories matching this topic...

SELGroup / SeSE

[UAI 2026 Oral] SeSE: Black-Box Uncertainty Quantification for Large Language Models Based on Structural Information Theory, which aims to detect hallucinated content in LLM-generated text.

python nlp black-box information-theory pytorch question-answering uncertainty-quantification trustworthy-ai llm hallucination-detection llm-hallucination semantic-entropy llm-uncertainty structural-entropy uai-2026

Updated Jul 6, 2026
Python

SuperMarioYL / probegate

Star

ProbeGate 是国产模型自治 agent 的 per-span 探针校验兜底门——把模型自报不确定度与机器可校验的 Autohand Code 探针配对，仅当双信号同时触发才路由人工。源自 Show HN「教会模型知道自己错在哪」命题，但拒绝信任黑盒自白。

agent calibration glm ai-safety guardrails abstention model-uncertainty qwen deepseek coding-agent human-handoff agent-guardrails ai-radar llm-uncertainty abstention-gate probe-validation

Updated Sep 14, 2026
Python

Add this topic to your repo

To associate your repository with the llm-uncertainty topic, visit your repo's landing page and select "manage topics."

Learn more