concept
active
concept:wait-we-don-t-need-to-wait-removing-thinking-tokens-improves-reasoning-efficiency-wang-et-al-2025aWait, we don't need to Wait! Removing thinking tokens improves reasoning efficiency (Wang et al., 2025a)
Baseline comparison method paper for reflection token suppression
Neighborhood — ranked by edge-count
Papers (1)
paper
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Filler tokens question.
- Mechanism for how the model modulates representation strength.
- Maximum token savings achieved by ReflCtrl on non-mathematical general reasoning tasks
- Explanation for the 'silent' thought phenomenon.
- Blocking attention to query token in layers 11–16 drops accuracy from 98% to 37%; with next token 8%finding0.734Attention knockout.
- Related work identifying neurons controlling short-thinking in reasoning models via weight editing
- Post-trained models repurpose user token activations to represent the Assistant's thoughts.hypothesis0.732Tested by J-lens comparison.
- Choosing token indices to sample adaptively, as in optimal experiment design, could be more efficient than post-hoc smoothing.hypothesis0.730Future work.