- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
This paper proposes SIVA-RL, a framework for assigning visual-intervention supervision according to observed changes in answer reward. A paired audit shows that different interventions frequently produce different correctness outcomes on the same example, motivating supervision at the sample–intervention level. SIVA-RL combines localized PatchSwap interventions, a clean-correctness gate, and reward-conditioned routing between sensitivity and invariance objectives. Training consists of a uniform-separation warm-up followed by a routed alignment stage. Across seven model–RL configurations, the paper reports improvements of 1.78–9.50 percentage points in average accuracy over the reported baselines.
Critical strengths
-
A fixed intervention can remove decisive evidence for one example while leaving another example largely unaffected. Adapting the auxiliary objective to the observed effect is a reasonable extension of methods that uniformly encourage separation.
-
The reported gains span multiple model sizes, three Qwen model families, and both GRPO and DAPO. Improvements occur in almost all individual benchmark comparisons, suggesting that the full training recipe has practical value.
-
The paper examines intervention construction, sensitivity routing, warm-up, invariance alignment, and shuffled routing. The invariance branch provides an additional improvement, and gains across different intervention operators suggest that the method’s value is not limited to the particular PatchSwap construction.
Critical weaknesses
-
Incremental contribution over related methods. Outcome-conditioned routing is the principal distinction from PAPO and DVRP, which already provide divergence-based perception supervision and sensitivity/robustness objectives, respectively. However, the comparison omits PAPO-DAPO-7B, whose published nine-benchmark average is 63.16, compared with SIVA-DAPO-7B’s 63.25 in Table 1. Cross-paper results cannot establish equivalence, but this close reference result makes matched comparisons with PAPO-DAPO and DVRP important for substantiating the incremental contribution.
-
Validation of the routing signal. The correspondence between reward changes and desirable alignment directions is not directly validated. Eq.7 and Eq.10 score both views against the original answer and route by reward drop. Hypothetically, task-preserving perturbations could induce failures through brittleness and trigger sensitivity supervision, while stable rewards could reflect shortcuts. These are potential failure modes, not demonstrated errors. The reported accuracy gains and divergence diagnostics do not quantify whether the selected directions correspond to the interventions’ semantic effects.
-
Isolation of directional routing. Table 4’s shuffled control supports the importance of sample–weight correspondence, but its interpretation as evidence for directional routing needs clarification. Eq.10 embeds the validity gate in the weights, and neither Table 4 nor Appendix specifies whether shuffling preserves eligibility. If eligibility changes, the 53.89-to-50.90 decrease may also reflect disrupted sample filtering.
Questions for authors
Please see the weaknesses.
Thank you for the clarifications and additional results. The semantic audit and eligibility-preserving shuffle resolve my concerns in W2 and W3. My remaining concern is the lack of comparisons with closely related methods, e.g. PAPO-DAPO, under matched training and evaluation settings. I recommend adding comparisons of more closely related methods to
Table 2to better substantiate the claimed benefit of outcome-conditioned routing over existing perception-aware objectives.Considering all these factors, I maintain my weak acceptance rating.