Research map / Large language models / Language, safety and interpretability
Fine-grained Uncertainty Decomposition in Large Language Models: A Spectral Approach
Nassim Walha, Sebastian G. Gruber, Thomas Decker, Yinchong Yang, Alireza Javanmardi, Eyke Hüllermeier, Florian Buettner
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace Input Clarification Ensembling with Spectral Uncertainty (aleatoric) | Phi-4 14B · AmbigQA | AUROC (%): 63.46 -> 69.15 | Improved |
| replace Input Clarification Ensembling with Spectral Uncertainty (aleatoric) | LLaMA 4 Maverick · AmbigQA | AUROC (%): 59.51 -> 60.39 | Improved |
| replace Predictive Kernel Entropy with Spectral Uncertainty (aleatoric) | Phi-4 14B · AmbigInst | AUROC (%): 75.93 -> 86.37 | Improved |
| replace Input Clarification Ensembling with Spectral Uncertainty (aleatoric) | LLaMA 4 Maverick · AmbigInst | AUROC (%): 69.66 -> 85.95 | Improved |
| replace Input Clarification Ensembling with Spectral Uncertainty (total) | Phi-4 14B · TriviaQA | AUROC (%): 89.45 -> 91.92 | Improved |
| replace Input Clarification Ensembling with Spectral Uncertainty (total) | LLaMA 4 Maverick · TriviaQA | AUROC (%): 82.76 -> 84.82 | Improved |
| replace Kernel Language Entropy with Spectral Uncertainty (total) | Phi-4 14B · Natural Questions | AUROC (%): 82.77 -> 81.63 | Hurt |
| replace Kernel Language Entropy with Spectral Uncertainty (total) | Phi-4 14B · Natural Questions | AUPR (%): 81.84 -> 81.98 | Improved |
| replace Input Clarification Ensembling with Spectral Uncertainty (total) | LLaMA 4 Maverick · Natural Questions | AUROC (%): 74.93 -> 75.02 | No clear effect |
Closest papers by meaning
- Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similarities (NeurIPS 2024 · similarity 0.92)
- Distinguishing the Knowable from the Unknowable with Language Models (ICML 2024 · similarity 0.92)
- Integrating Local and Global Entropy for Uncertainty Quantification in LLMs (NeurIPS 2026 · similarity 0.91)
- Inv-Entropy: A Fully Probabilistic Framework for Uncertainty Quantification in Language Models (NeurIPS 2025 · similarity 0.91)
- Decomposing Uncertainty for Large Language Models through Input Clarification Ensembling (ICML 2024 · similarity 0.91)
- Graph-based Uncertainty Metrics for Long-form Language Model Generations (NeurIPS 2024 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
