- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper proposes an inference-time attack that increases demographic bias in a diffusion language model. It constructs a direction from the mean activation difference between target and comparator answers, then uses a proportional–integral controller to adjust its strength from the target-answer probability during denoising. The main contribution is feedback across denoising steps, together with an empirical comparison against fixed-strength interventions. On ambiguous BBQ questions with LLaDA-8B (Black target), the target–comparator gap rises from 1.8 to 16.7 pp, compared with 3.5 for a constant mean strength. On SocialStigmaQA, stigmatizing answers rise from 17.6% to 58.1%. The effect transfers to Arab and Old targets. It fails on Asian, Latino and White because of high invalid-output rates, the effect is small on Dream, and it is beaten by constant steering on LLaDA-MoE. The authors argue that dLLMs expose a new attack surface because the answer can be observed and steered before it is committed.
Critical strengths
- The paper identifies a useful inference-time control opportunity in dLLMs: repeated predictions at an unresolved answer position can provide feedback for adjusting an intervention before commitment. The primary LLaDA results show substantial demographic target-answer shifts. The matched comparisons hold the steering direction and injection sites fixed, helping isolate the contribution of how the strength is set.
- The evaluation includes several valuable checks: answer-order rotations, matched-strength and per-example constant baselines, item-clustered confidence intervals, and parser-sensitivity analyses. The authors also report unsuccessful demographic targets and models where feedback does not outperform constant steering (Asian, Latino, White, LLaDA-MoE). These checks and negative results help establish the scope of the empirical finding.
Critical weaknesses
- The experiments don’t show that feedback is what drives the effect. All the ablations compare PI against a strength that stays constant across denoising steps. But PI’s strength increases over the steps (Fig. 1c). So a fixed schedule that increases over time, with no feedback, could potentially explain much of the gain. The sampling setup makes this harder to judge. Only the first token is scored, but the model generates 32 tokens over 64 steps with steering at every position, the last 32 steps commit nothing, and the step at which the answer token commits is never measured. That step determines how many controller updates actually matter. Finally, the criteria for a successful attack change between settings. Asian, Latino and White runs are discarded for high strict invalidity, but SocialStigmaQA PI is 17.4% strict-invalid (Table 2), above the paper’s own 15% cutoff (Sec. 4.4). That result relies on a semantic parser added after inspecting the outputs.
- The threat model needs more motivation. The attacker can hook activations and read the sampler’s answer probabilities. In this setting, though, the same access would usually allow editing the logits or the sampler directly, for example by adding a bias to the target letter’s logit. The paper excludes these by assumption, but it doesn’t explain why an attacker would choose activation steering instead. A clearer argument for when this attacker is realistic, or framing the work as a study of how steerable dLLMs are during denoising, would make the contribution easier to judge.
Questions for authors
Clarifications:
- When does the answer token typically commit under your attack, and how many controller updates come before that? Prior work [1] finds that LLaDA’s answers on multiple-choice tasks are usually settled by halfway through denoising. Related to this: why generate 32 tokens over 64 steps when only the first token is scored, and does steering every position raise the invalid rate? I don’t expect new runs here, but please explain these sampling choices.
- How was block 14 chosen for extracting the direction, and why is it then added at every block and position? Was any ablation done over these choices? Why does the attack fail on the Asian, Latino and White targets: is it the direction itself, the strength, where it is added, or some bias inherent in the model?
- The threat model allows activation hooks and access to intermediate answer probabilities but excludes direct logit and sampler edits. Can you clarify what deployment scenario motivates this particular access boundary? Addressing this would help establish the practical significance of the attack beyond demonstrating controllability during denoising.
- Can you please compare PI with a predetermined time-varying schedule without feedback?
On SocialStigmaQA:
- What is the unsteered gap under the semantic parser? Reporting it with the semantic-parser outcome rates would make the magnitude of easier to interpret.
- Why isn’t the semantic parser also applied to Table 1?
Suggestions (do not affect my score):
- The table formats are inconsistent: Table 2 drops the Gap column and Table 3 drops Tgt and Cmp. A single format would make them easier to compare.
- Line 345 says no baseline exceeds 4.6, but the per-item static strengths reach 4.8 and 8.4. Please clarify which rows count as baselines.
References
[1] P. Li et al. Diffusion Language Models Know the Answer Before Decoding. arXiv:2508.19982, 2025.
11 comments
1VHG-Q2. Why extract at layer 14 and inject at every block? What explains the failed targets?
The gap persists across the tested extraction layers; injection coverage matters substantially. Section 3 specifies extraction at layer 14 and injection at every block and position. We additionally varied extraction depth while retaining all-block injection, then tested injection at block 14 alone (100 items × 3 rotations).
| Direction extraction / injection | Gap (pp) | Invalid (%) |
|---|---|---|
| Layer 14 / all blocks (paper setting) | 19.3 | 11.0 |
| Layer 8 / all blocks | 18.7 | 18.3 |
| Layer 20 / all blocks | 20.0 | 23.0 |
| Layer 26 / all blocks | 16.3 | 6.0 |
| Layer 14 / block 14 only | 2.3 | 0.0 |
All tested extraction layers produce gaps of 16 to 20 pp, although validity varies. Layer 14 gives 19.3 pp with 11.0% invalidity; layer 26 gives 16.3 pp with 6.0% invalidity. Block-14-only injection gives 2.3 pp, supporting the all-block design under these settings without claiming it is the only viable configuration.
The failed targets show a mismatch between the letter sensor and the generated output. Section 4.4 reports Asian, Latino and White as failures. Additional output inspection finds “Li…” in 1,064 of 1,200 Asian prompts, surname fragments such as “Z:” and “…PEZ” for Latino, and repetition (“AAAA…”) for White. Target-letter probability falls below base: Asian 0.044 vs 0.189; White 0.031 vs 0.196. These observations suggest that the intervention can promote names or malformed text rather than the measured answer letter, causing the controller to increase strength as the letter-probability shortfall persists. They do not isolate direction quality from controller and injection effects.
| Target | Base gap (pp) | PI gap (pp) | PI invalid (%) | Dominant invalid output |
|---|---|---|---|---|
| Black | 1.8 | 16.7 | 7.8 | Mixed ("Answer:", "The …") |
| Arab | −0.2 | 22.3 | 9.7 | Empty output |
| Old | −5.5 | 31.8 | 9.3 | Empty / "The …" |
| Latino | 4.2 | 15.2 | 59.6 | Surname fragments ("Z:", "…PEZ") |
| Asian | −3.2 | 5.6 | 93.8 | "Li…" |
| White | −1.6 | 3.9 | 96.1 | "AAAA…" |
Black, Arab and Old retain substantially better validity. The additional Latino suppression result (SsMd-W1) also shows that this direction’s usefulness depends on the intervention setting, rather than establishing that it is uniformly ineffective.
1VHG-W2 / Q3. Why study activation steering when logit or sampler edits may be available?
Our question is what feedback can achieve through activation steering under the access boundary in Section 2. We do not claim that steering is stronger than unrestricted logit or sampler editing. Sections 2 and 5 cite activation-access precedents: Trojan Activation Attack (Wang & Shu, CIKM 2024) studies inference-time steering, including bias; Arditi et al. (NeurIPS 2024) removes refusal through a representation direction; Schwinn et al. (NeurIPS 2024) studies continuous-space attacks; and Shnaidman et al. (ReALM-GEN at ICLR 2026) applies activation interventions during LLaDA denoising.
Our assumed interface explicitly includes intermediate probability readout and bounded residual-stream intervention, while excluding direct logit and sampler edits. These precedents motivate the interface but do not establish that every compromised serving process is restricted to it. We will clarify that boundary.
The additional open-ended experiment demonstrates that the attack is not confined to MCQ answer letters. With one-sentence BBQ responses, the same actuator and a target-name probability sensor yield +28 pp [8, 48], the only tested method whose reported shift interval excludes zero (R1.1). The MCQ implementation does use the target’s answer letter for feedback; the free-form result supports broader controllability, not a claim that logit editing cannot operate on free text.
1VHG-Q5. What is the clean SocialStigmaQA gap under semantic scoring?
The clean gap is −44.3 pp under both parsers, with 0% invalidity. Table 2 reports the strict baseline and parser-dependent shifts; Appendix C.8 defines the semantic rules. We now make the absolute gaps explicit below. The run generated 1,665 prompts; as specified in Appendix C.8, headline race metrics use the race subset, while no-stigma prompts are diagnostic.
| Condition (1,665 prompts) | Gap, strict (pp) | Invalid, strict (%) | Gap, semantic (pp) | Invalid, semantic (%) |
|---|---|---|---|---|
| Clean | −44.3 | 0.0 | −44.3 | 0.0 |
| Decode-PI | +39.5 | 17.4 | +55.6 | 1.4 |
| Open loop α = 4 | +18.3 | 3.9 | +19.4 | 37.8 |
| Open loop, effort-matched | −2.5 | 0.0 | −2.5 | 0.0 |
Decode-PI shifts the gap by +83.8 pp under strict parsing and +99.9 pp under semantic parsing, with semantic invalidity 1.4%. Against the effort-matched open loop, paired gains are +42.0 pp [36.2, 47.9] and +58.1 pp [48.8, 67.7], respectively. Thus, the advantage holds under both scoring rules.
1VHG-Q6. Why was semantic scoring not applied to BBQ Table 1?
The additional rescoring shows that the main BBQ comparison is stable under both parsers. Appendix C.8 introduced semantic scoring to handle exact answer text and reject option re-listings on SocialStigmaQA. We have now applied both parsers to the Table 1 conditions:
| Condition (1,200 prompts) | Gap, strict (pp) | Invalid, strict (%) | Gap, semantic (pp) | Invalid, semantic (%) |
|---|---|---|---|---|
| Base | 1.8 | 0.0 | 1.8 | 0.0 |
| Decode-PI | 16.7 | 7.8 | 16.8 | 11.1 |
| Constant α = 3.28 | 3.5 | 20.8 | 3.5 | 20.8 |
| Per-item static α (oracle) | 4.8 | 13.1 | 4.8 | 13.1 |
| Per-item static α (hindsight) | 8.4 | 13.4 | 8.3 | 13.5 |
| CAA L14 (a = 16) | 3.3 | 0.0 | 3.3 | 0.0 |
| ActAdd L14 (a = 16) | 4.1 | 0.0 | 4.1 | 0.0 |
| Mean-AcT (s = 2) | 3.5 | 0.2 | 3.5 | 0.2 |
| Linear-AcT (s = 1) | 2.6 | 0.1 | 2.6 | 0.1 |
| AurA inject (g = 4) | 0.0 | 0.2 | 0.0 | 0.2 |
| AurA vanilla | 0.8 | 0.0 | 0.8 | 0.0 |
| ITI-C (K = 48, a = 8) | 2.1 | 0.0 | 2.1 | 0.0 |
No reported gap changes by more than 0.2 pp. Decode-PI remains at 16.7–16.8 pp, above the listed steering baselines and static controls. Its invalidity increases from 7.8% to 11.1%, so parsing does affect validity, but PI remains below the 15% threshold under both rules. We will report both.
1VHG-W1. Is the SocialStigmaQA success criterion inconsistent with the 15% invalidity threshold?
Yes, SocialStigmaQA PI exceeds the 15% threshold under strict parsing. Section 4.4 and Table 2 report 17.4% strict invalidity; Appendix C.8 identifies semantic parsing as a secondary analysis added after output inspection. We will distinguish strict-threshold failure from semantic-parser success explicitly.
The additional audit finds that 250 of the 271 strict-invalid outputs are the bare biased answer word (“Yes” or “No”), rather than an answer letter. Under the semantic rules applied to every condition, PI’s invalidity falls to 1.4% and its gap increases from +39.5 to +55.6 pp. This supports a substantive answer shift despite letter-format violations, without retroactively treating the strict criterion as satisfied. We will report both parsers and their rules consistently.
1VHG. Presentation and reporting suggestions
We will standardize Tables 1–3 to include target, comparator, gap and invalidity, identify each parser, and clarify the denoising schedule. We will also replace “no baseline exceeds 4.6 pp” with a precise comparison: the tuned open loop reaches 4.6 pp, the probe-based per-item static control 4.8 pp, the hindsight static control 8.4 pp, and Decode-PI 16.7 pp. Only the hindsight condition uses PI’s realized trajectory to set its constant.
The new schedule controls directly address the reviewer’s central alternative: predetermined time variation, including replayed PI trajectories, does not recover the advantage of feedback tied to the current example’s predictions.
Response to Reviewer 1VHG
We thank the reviewer for recognizing the denoising-time attack surface and matched comparisons . We clarify the existing evidence and report the additional experiments below.
1VHG-W1 / Q4. Does feedback matter beyond a predetermined time-varying schedule?
Yes. The additional schedule controls support feedback as the source of the advantage. Section 4.3 and Table 1 compare PI with constant and per-item static strengths. To test the reviewer’s time-varying alternative directly, we replaced feedback with PI’s average α(t) curve, another item’s actual PI trajectory, and a linear ramp from 0 to 6.5. The three schedules have reported mean α = 3.27 over the committing steps. The new conditions were run together on 1,200 prompts; the paper’s constant-strength result is included separately for reference.
PI exceeds all three no-feedback schedules, with paired advantages of +13.4 [9.2, 17.6], +12.3 [8.5, 16.2], and +13.5 [9.2, 17.7] pp. Replaying another item’s trajectory preserves the replayed command sequence but removes its dependence on the current item’s predictions. Its failure to recover PI’s gap directly supports the value of closing the loop, beyond increasing strength over time. We will add these controls and distinguish this rerun from Table 1.
1VHG-Q1. When does the answer commit, and why use 32 tokens, 64 steps and sequence-wide steering?
Early predictability does not end the intervention window: commitment does. Section 3 describes the 32-token, 64-step schedule. Our additional logging distinguishes a stable answer-slot argmax from the sampler’s commitment decision. With gen_length = block_length = 32, one token commits in each of steps 0–31; steps 32–63 commit nothing. SocialStigmaQA’s 32-step and 64-step runs produce identical outputs on all 1,665 prompts.
For unsteered prompts, the answer argmax settles at median step 0, consistent with the reviewer’s observation. However, commitment occurs at median step 31, with the answer committed last in 74% of prompts. On the 223 prompts PI flips to the target, the answer settles at median step 29 (97% after halfway) and commits at step 31. Thus, feedback can change the answer before commitment even when the unsteered prediction stabilizes early.
The additional ablations support the chosen configuration: answer-position-only steering gives 1.9 pp, while 8 tokens / 8 steps gives −2.3 pp under PI. Neither recovers the 15.1 pp sequence-wide result. The shorter configuration changes both generation length and the feedback horizon, so it does not isolate update count alone. Answer-only steering reduces invalidity to 0.0% but also removes the observed gap increase.