- Rating
- 1
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper tests whether choosing the domain-matched private expert in a shared-private SAE preserves downstream predictions better than choosing the other expert. It estimates shared structure from principal-vector contrasts and edits only the PCA-modeled component. Experiments cover GPT-2, Pythia-70M, and cell embeddings, with additional simulations and an independent text evaluation.
Critical strengths
-
The intervention preserves the unmodeled residual exactly. Paired routing comparisons and shared-only and randomized-route controls make the downstream effect of switching fitted experts easy to interpret.
-
The paper reports important qualifications. The Pythia experiment without a forced subspace intersection retains positive effects, while the independent GPT-2 replication does not. The RxRx1 results provide evidence beyond language models.
Critical weaknesses
-
The routing comparison does not separate functional ownership from ordinary reconstruction quality. Each private expert is fitted to its own domain, so the wrong expert may simply reconstruct that domain less accurately. Preserving the orthogonal residual does not control this difference inside the edited span. Sections 4 and 6 promise edit magnitudes and output divergence, but the corresponding numerical comparisons are absent. Correct routing itself adds about 2.47 NLL in the independent GPT-2 index-3 example. The routing effect supports relative downstream preservation, but the missing diagnostics leave the introduction's distinction from better fitting domain statistics unresolved.
-
The paper explicitly disclaims estimator superiority and semantic identifiability in Section 6, but the practical value of its ownership cutoff still has limited validation. Appendix A.3 shows that horseshoe ordering is identical to raw contrast-energy ordering, leaving the rank cutoff as the consequential choice. The fixed noise and slab variances lack sensitivity evidence, and the primary design's shared rank is largely constrained by the forced intersection. The additional experiment establishes a routing effect without that intersection, but does not establish accurate ownership recovery. Appendix F mainly reports downstream responses rather than recovery of known ownership by the full estimator.
Questions for authors
-
Can you report reconstruction error, edit RMS, KL divergence, and absolute NLL for the original, correct-route, and wrong-route conditions? Does the routing advantage persist when reconstruction damage is matched?
-
How sensitive are selected ranks and route effects to the noise and slab variances? How does the estimator compare with a simple principal-angle cutoff or matched-rank domain PCA on the same data?
Response to Authors
I would like to thank the authors for their detailed response. I have reviewed their responses to all three reviews, as well as the general response.
It is helpful to know that "shared-private SAEs" are indeed intended to be a contribution of this work, and I appreciate the results of some of the suggested followup experiments. However, the main problem remains: what is the purpose of this study? The authors responded that:
And in responding to reviewer mVfc, they state that:
Why does this matter? It seems that the question is inherently restricted to a new class of SAEs that the authors introduce in this work (shared-private SAEs) -- in effect, the goal seems to be to define a new architecture and answer a specific question about it; but neither the architecture nor the question has clear motivation. The architecture is designed around the question about private vs shared vectors, and the question of private vs shared vectors is explored with respect to the only architecture that (to my knowledge) has this specific feature (decomposition into shared vs private domains). The motivation is circular: the architecture only matters because of the question, and the question only matters because of the architecture.
While I appreciate the authors' effort to address some technical questions and run followup experiments, the core weakness remains, so I don't see any reason to change my rating.