Fetching from the wire…
01
02
03
04
05
06
07
08
Research2026-08-12 · source-backed
The argument is that token uncertainty shows up not only in output-distribution breadth but in whether a confident prediction is fragile under perturbation of its attention pathways (arXiv 2608.11138). It's training-free: mask attention heads, measure BALD mutual information among resulting subnetworks through a semantic-agreement kernel. Single-response variant ties or beats Semantic Entropy on 10 of 12 grounded benchmark-backbone settings. The honest limit the authors state plainly: on parametric QA all variants revert to or below the zero-cost MSP baseline. This helps for retrieval-grounded answers, not closed-book ones.
Each link below shares sources, entities, or timing with this story.
Same source domain / Shared topic / Downstream implication
Reported by the same outlet (arxiv.org); overlapping topics (answer, argument, author); traces where this leads (downstream).
Same source domain / Shared topic / Tension
Reported by the same outlet (arxiv.org); overlapping topics (author, beat, variant); pushes against this story (but).
Reported by the same outlet (arxiv.org); overlapping topics (baseline, beat); pushes against this story (against).
Reported by the same outlet (arxiv.org); overlapping topics (answer, attention); pushes against this story (versus).
Reported by the same outlet (arxiv.org); overlapping topics (baseline, beat); pushes against this story (against).
Reported by the same outlet (arxiv.org); overlapping topics (author, beat); pushes against this story (against).
Reported by the same outlet (arxiv.org); overlapping topics (baseline, beat); pushes against this story (versus).
Same source domain / Shared topic
Reported by the same outlet (arxiv.org); overlapping topics (baseline, beat, prediction).