Decoder-only transformers fail to base verification decisions solely on current search state in cumulative traces because of scattered retrieval and history entanglement; Selective State Attention enforces state-only decisions via a fixed mask.
Advances in Neural Information Processing Systems , volume=
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3representative citing papers
Distillation signals align better with ideal updates on incorrect student rollouts than correct ones, with optimal teacher context depending on student capacity and task.
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核
citing papers explorer
-
Can Transformers Learn to Verify During Backtracking Search?
Decoder-only transformers fail to base verification decisions solely on current search state in cumulative traces because of scattered retrieval and history entanglement; Selective State Attention enforces state-only decisions via a fixed mask.
-
Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why
Distillation signals align better with ideal updates on incorrect student rollouts than correct ones, with optimal teacher context depending on student capacity and task.
-
Mechanisms of Misgeneralization in Physical Sequence Modeling
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核