- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
TrajEdit is a framework for editing unsafe behavior in language models through sparse, multiple-anchor constraints. Instead of supervising an entire safe response, it selects the first response chunk and additional high-perplexity chunks from an unsafe generation. For each selected anchor, it constructs a safe continuation that shares the preceding harmful prefix, then applies an existing editing solver at the corresponding divergence point.
Critical strengths
- Empirical support for localization. The annotated harmful-onset evaluation goes beyond illustrative examples and reports quantitative evidence that perplexity carries localization information.
- Substantial capability preservation relative to the selected baseline.
Critical weaknesses
- The sequential evaluation is insufficiently specified. It is unclear which model checkpoint generates each reported measurement, how previous edits are retested, and whether generalization cases remain disjoint from later editing inputs.
- Capability preservation is incomplete and narrowly evaluated. Table 4 evaluates one model and one baseline configuration. TrajEdit still reduces MMLU from 0.4125 to 0.3345, an absolute drop of 7.8 percentage points.
- Uncertainty and cost reporting are incomplete. Results lack confidence intervals and variability across edit orders or random truncation choices. Editing-time units and whether timing includes anchor scoring, safe-target generation, and discriminator calls should also be specified.
Questions for authors
See weakness