ToxiREX is a new dataset of 128k Reddit comments in six languages with hierarchical annotations for implicit toxicity in conversational context based on an existing reasoning schema.
From memorization to reasoning in the spectrum of loss curvature
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Contribution Weights combine attention, value magnitude, and directional alignment to measure token influence more faithfully than attention alone, and show attention sinks actively suppress information via a convex sink-rate to output-norm relationship.
citing papers explorer
-
ToxiREX: A Dataset on Toxic REasoning in ConteXt
ToxiREX is a new dataset of 128k Reddit comments in six languages with hierarchical annotations for implicit toxicity in conversational context based on an existing reasoning schema.
-
Contribution Weights: A Geometrical Analysis of Self-Attention Transformers
Contribution Weights combine attention, value magnitude, and directional alignment to measure token influence more faithfully than attention alone, and show attention sinks actively suppress information via a convex sink-rate to output-norm relationship.