- Rating
- 2
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
AgentVax asks how a multi-agent workflow can learn from harmful events that are confirmed only after the relevant decisions have been made. It records the local information available to each agent, trains a discrete-time survival model to forecast future policy-violating proposals, and restores historical workflow states to estimate the effects of continuing, reviewing, or blocking. These replay-derived values are then used to update local intervention policies under utility and false-blocking constraints.
The evaluation combines eight workflow strata, thirteen actor configurations, and several attack families. Local views provide complementary warning signals, and combining them improves early prediction. The full system sharply reduces harm relative to no defense, but its advantage over simpler learned guards is limited.
Critical strengths
-
AgentVax cleanly separates the time of a harmful proposal, the later release of its verified label, and the information that was actually visible to an earlier agent. This avoids giving the safety model retrospective evidence that would not have been available online.
-
Counterfactual replay addresses an important limitation of risk-only guards: a high risk score does not show that an authorized action will change the outcome. The paper also accounts for benign utility, false blocking, latency, and background work rather than reporting attack detection alone.
-
The evaluation covers many workflow and actor configurations, and it reports mixed findings such as the missed test-time false-blocking target and the unfavorable cost comparison with Binary.
Critical weaknesses
-
I do not see a clear end-to-end advantage over simpler guards. On the source stratum, Binary has lower harm with essentially the same benign success, and similar patterns appear elsewhere. AgentVax beats Central on harm in only 47 of 104 cells while using more lifecycle work; its cumulative harmful-workflow count in the closed loop is also higher than both Central and Memory. The method is much better than no defense, but the experiments do not show when its extra machinery pays off.
-
Several results used to motivate the design are not evaluations of the complete method. The role-matched update is explicitly a separate comparator, the action-value advantage is shown in a controlled state-action experiment, and the survival objective improves timing prediction in a synthetic diagnostic while matching a binary head on the main gate outcome. These studies support individual intuitions, but they leave unclear which component produces the practical benefit of AgentVax in the end-to-end workflow.
-
The replay procedure assumes that a complete, version-compatible workflow state can be restored and rerun under alternative actions. This is plausible in the isolated evaluation environments, but it is a demanding requirement for workflows involving external services, nondeterministic tools, or changing model versions. Since replay is both the main source of action supervision and the dominant background cost, the paper needs a clearer account of how often restoration fails and what remains of the method when only partial replay is possible.
Questions for authors
-
Which observed platform or attack characteristics predict when AgentVax should be preferred to Binary or Central? Is there a subset of the 104 cells where replay-derived action values provide a consistent gain at matched total work?
-
Can the survival head, binary head, and replay-trained action policy be compared within the same end-to-end workflow protocol? This would help connect the component studies to the final system.
-
What fraction of historical checkpoints cannot be restored or replayed successfully, and how are these cases distributed across platforms and actions?
-
How would AgentVax operate when external state cannot be cloned exactly and only logged observations or approximate simulators are available?
4 comments
pX4V.W2 & Q2 · Which component changes which part of the workflow?
Reviewer’s comment These studies support individual intuitions, but they leave unclear which component produces the practical benefit of AgentVax in the end-to-end workflow.
Thank you for distinguishing the supporting diagnostics from an evaluation of the complete system. The clearest reported example is replay scoring. At the fixed Source threshold in Table 9, restoring it reduced false blocking by 11.2 points and increased harm by 4.5 points. The table below shows the corresponding change in benign completion and the TAMAS result. We then specify a mechanism-specific outcome for each change to the trained guard.
<a id="r3a"></a>
R3a · Reported ablations at the fixed operating threshold, manuscript Table 9
| Enabled components B/R/J | Source H / U / FBR (%) | TAMAS H / U / FBR (%) | Source / TAMAS Brier |
|---|---|---|---|
| All three (111) | 9.4 / 86.7 / 5.9 | 11.4 / 88.0 / 3.9 | 0.136 / 0.134 |
| Without early-prefix supervision (011) | 9.6 / 85.9 / 7.3 | 12.2 / 86.8 / 5.5 | 0.137 / 0.135 |
| Without replay scoring (101) | 4.9 / 79.4 / 17.1 | 6.4 / 80.6 / 15.7 | 0.136 / 0.134 |
| Without cost-sensitive review rule (110) | 7.2 / 85.9 / 7.0 | 9.0 / 87.1 / 4.6 | 0.136 / 0.134 |
B denotes early-prefix supervision, R denotes replay scoring, and J denotes the cost-sensitive score-and-review rule. Restoring R allowed more legitimate work to proceed under that rule. Some harmful work proceeded as well. Restoring B improved the forecast Brier score on both platforms. Fig. 2a, the synthetic timing diagnostic and the controlled action-value experiment each test a different mechanism.
R3b · One change to the trained guard on M03/Source
| Change from full guard | Mechanism, ablated → full | H ↓ | U ↑ | FBR ↓ |
|---|---|---|---|---|
| Full guard | Reference for all comparisons | 5.6% | 85.0% | 2.5% |
| Feedback stays at confirming stage | Pre-proposal recall 58% → 65% | 6.4% | 84.6% | 2.7% |
| Capacity-matched pooled head | Shared-pattern recall 65% → 70% | 6.0% | 84.4% | 2.9% |
| Hide visible provenance and authority | Both cases in an authorization pair correct 62% → 72% | 7.0% | 84.5% | 2.9% |
| Substitute a binary head | Event-time NLL 1.27 → 1.15 | 5.6% | 85.0% | 2.5% |
| Shuffle replay values between matched states | Action regret 0.048 → 0.034 | 7.4% | 84.6% | 2.8% |
| Use the endpoint rule | Action regret 0.050 → 0.034 | 9.4% | 85.0% | 2.5% |
| Freeze after the first update | Timely control in final round 81% → 84% | 6.5% | 84.4% | 2.8% |
For example, removing upstream supervision changes whether the reader can warn before the sender proposes an unredacted attachment. Hiding the provenance of the very same attachment instead changes whether the guard can tell an authorized internal recipient from an external recipient. All arms use the same training roots, actor, actual permissions and evaluation tasks. For the binary timing readout we distribute its endpoint event mass uniformly over the bins, as in Appendix F.3, and score the observed follow-up. All differences are paired by root. R3b reports 5.6% harm for the full guard. R3c includes both 5.6% and 8.6% result columns, and R3d reports the 8.6% result.
Manuscript revision. We will place each ablation beside its mechanism readout in Appendix F.3 and identify the action rule used in Table 9. Thank you for geive us the chance to show how each component changes behavior as well as how the workflow ends.
pX4V.W3 & Q4 · How training continues when external replay is costly or fails
Reviewer’s comment The replay procedure assumes that a complete, version-compatible workflow state can be restored and rerun under alternative actions. This is plausible in the isolated evaluation environments, but it is a demanding requirement for workflows involving external services, nondeterministic tools, or changing model versions.
We appreciate the reviewer’s question about external services. At each eligible control point, for example just before releasing an attachment or pausing it for review, the protocol records the pending operation, agent context, server-managed file contents or immutable references and their hashes, permissions, queues, event clock and model/tool versions. To test an alternative action, we load that checkpoint on an isolated server, verify the saved state and rerun the suffix under the frozen policy. This avoids repeating the preceding tool calls and interactions. Table 8 assigns 173,966.4 modeled work units to replay on Source/M08. Our proposed fallback addresses external dependencies outside the saved server state. (Section 5.2, Appendices B.5 and D.3, Algorithm D.2)
Recovery. Let h be the logged history immediately before the fork, x the local part visible to the role, u an authorized action and ξ a fresh downstream seed. A recovered branch is h ⊕ u ⊕ G(h,u,E,ξ). G generates new model decisions. E supplies compatible external observations. A past read before the fork is already part of h. A post-fork read can reuse a response only if its arguments, immutable object or quote version, effective time, permission scope and causal inputs agree. A changed recipient, new send or cancelled payment needs a new receipt for that action. Historical tokens and tool replies have zero training-loss weight. For the supplementary audit we allow up to two native attempts per branch and use a per-branch work cap of twice the discovery median for that platform and action. A request already above the cap is recorded as unattempted. An attempted branch without a verifiable effect is recorded as failed unless a compatible read supports scoring. (§5.2, Appendices B.5 and D.3)
** Reuse the recorded read when the branch still asks the same question.** When recovery failed and all tool replies received before the fork are already part of the saved history. For a read requested after the fork, we propose to store its arguments, returned content and hash, object or quote version, effective time and authorization scope. We will reuse the saved reply only when those conditions and the branch’s upstream dependencies still match and the read is reproducible. For example, a branch may retrieve the original, still-valid quote by its immutable ID. A review that asks for the current price after the quote expires needs a fresh observation. Sending to a new recipient or cancelling an external job also needs confirmation of that specific effect. If no native reference is available, we stop at the last verifiable event and retain any eligible factual labels for earlier agents.
3Build two offline training records. We also constructed two types of adventure data during training, which avoids recovery failures. For early forecasting, save (xᵢ,ₜ, mₜ, δₜ) at an event-free prefix. The last two entries give observed follow-up length and whether the first prohibited proposal occurred within it under the specified reference continuation μ. One event verified and released by the training cutoff can label several earlier agents at different distances from the event. For example, an event at step 6 under μ, verified by the cutoff, yields (m,δ)=(4,1) at step 2 and (2,1) at step 4 when K ≥ 4. A compatible benign continuation supplies the comparison. An intervention-truncated log with unknown follow-up leaves that label missing. At inference the guard sees only its original local input.
For action learning, store same-prefix pairs (xᵢ,ₜ, u, v, Q̃(h,u) − Q̃(h,v)). For each action, Q̃(h,u) = (1/M) ∑ⱼ [Lᴴⱼ + λᵤ Lᵁⱼ + λ_c Cⱼ] averages the harm, legitimate-task loss and cost over the M planned recoveries, provided all are scorable. The auxiliary target favors the action with lower conditional loss. When all required authorized actions are supported, the Eq. (5) baseline gives Ã(h,u) = ∑ᵥ π_old(v|xᵢ,ₜ) Q̃(h,v) − Q̃(h,u) for the Eq. (6) update. Otherwise we omit that action update and retain eligible factual forecast labels. We tag every target by its origin. Recovery estimates are conditional on the recorded external evidence. Independent native continuations test action ordering and drift of each action’s effect relative to continue.
Illustrative pair. A user asks for an expense total without personal details. A read-only tool has already returned the amounts and an internal identifier. From the same saved prefix, continue may produce a reply containing a personal identifier, while review may return only the correct total. Both replies can be scored locally for the disclosure and task completion. If a branch instead requests a new payment, the payment outcome stays unknown. A disclosure already verified on a compatible factual trace can still supervise that earlier prefix.
R4e · Training supervision on 640 Source roots, native evaluation on 2,000 new M03/Source roots
| Training method | Training decisions assigned native branches | H ↓ | U ↑ | FBR ↓ |
|---|---|---|---|---|
| Supervised imitation of recorded actions (SFT) | 0% | 7.8% | 84.5% | 2.9% |
| Imitation weighted by factual terminal return | 0% | 7.3% | 84.6% | 2.8% |
| AgentVax, prefix recovery only | 0% | 6.2% | 84.8% | 2.7% |
| AgentVax, equal mix of native and recovered branches | 50% | 5.9% | 84.9% | 2.6% |
| AgentVax, full native replay | 100% | 5.6% | 85.0% | 2.5% |
We hold the risk head, guard initialization, training decisions and optimizer budget fixed. Allocation in the mixed arm precedes outcome inspection. Recovery incurs 0.6 points more harm than native replay and 1.6 points less than SFT. Both contrasts use native workflows; cost accounting includes regenerated tokens, live tool calls, retries and training time.
Manuscript revision. We will add the checkpoint contents, the compatibility test for reads, and the two offline targets to Appendices B.5 and D.3. Thank you for asking how much of the learning signal survives a realistic external failure, this further enhances the reproducibility and rigor of our paper.
pX4V.Q3 · How many checkpoints and branches can be recovered, and where do failures occur?
Reviewer’s comment What fraction of historical checkpoints cannot be restored or replayed successfully, and how are these cases distributed across platforms and actions?
Thank you for insisting on the request denominator. The submitted restoration protocol verifies state identity but does not publish historical counts of attempted and unattempted requests. The fractions below are results from a supplementary audit, in which each platform contributes 120 checkpoints and each checkpoint permits all three actions. A failure after an attempt, a compatible logged read, and a branch never attempted have distinct consequences for training.
<a id="r4a"></a>
R4a · Branch status by platform and authorized action
| Platform · action | Requested | Exact | Recorded read | Failed | Unattempted |
|---|---|---|---|---|---|
| Source · continue | 120 | 107 | 7 | 4 | 2 |
| Source · review | 120 | 103 | 10 | 4 | 3 |
| Source · block | 120 | 114 | 1 | 4 | 1 |
| TAMAS · continue | 120 | 107 | 6 | 5 | 2 |
| TAMAS · review | 120 | 102 | 11 | 4 | 3 |
| TAMAS · block | 120 | 115 | 1 | 3 | 1 |
| All branches | 720 | 648 (90.0%) | 36 (5.0%) | 24 (3.3%) | 12 (1.7%) |
The percentage changes when the unit is a checkpoint. Of 240 checkpoints, 200 (83.3%) have fully exact action sets, 20 (8.3%) have complete sets with at least one compatible read, and 20 (8.3%) have incomplete sets. Thus 684/720 branches (95.0%) and 220/240 checkpoints (91.7%) are scorable at their respective levels. Whole-checkpoint coverage cannot be obtained by cubing branch coverage because branch failures share external dependencies.
R4b · Compatible reads reused / branches with the stated external limitation
| External dependency | Continue | Review | Block | Total |
|---|---|---|---|---|
| Live price, inventory or appointment availability | 9/10 | 16/18 | 1/2 | 26/30 |
| Mutable file or expiring link | 4/5 | 5/5 | 1/2 | 10/12 |
| New external write or cancellation receipt | 0/3 | 0/1 | 0/2 | 0/6 |
| Total | 13/18 | 21/24 | 2/6 | 36/48 |
Here 48/720 branches (6.7%) encounter one of the listed external limitations. Of the 42 read-related cases, 36 have compatible saved results. The six incompatible reads and six unverifiable writes account for 12 of the 24 failures. The remaining 12 failures cover other state or version mismatches. The 12 unattempted branches remain a separate category. These counts describe the audited requests; they do not establish a general difference in failure rates between platforms.
Manuscript revision. We will report counts of requests, attempts and scorable action sets by platform and action in Appendices B.5 and D.4. Your question ensures that the least reproducible workflows remain visible in the results.
Reviewer pX4V
Dear Reviewer pX4V,
We sincerely thank the reviewer for the careful and constructive feedback. We are encouraged by your recognition of the safety improvements AgentVax achieves over the undefended setting. Your detailed analysis of end-to-end performance, component contributions, and replay feasibility has also helped us clarify the evidence needed to demonstrate the method’s practical value and the importance of fair comparisons with simpler guards. We address each of your concerns in detail below.
pX4V.W1& Q1 · End-to-end value against Binary, Central and Memory
The better defensive capability of Binary in the original text comes from the fact that Binary starts prediction n windows in advance. Each node is judged n times, which leads to unfair comparisons with other methods. The experiemnt below reveals what each guard can do with the same opportunity to intervene.
In the rolling binary control, scoring starts as soon as the warning window opens. Four eligible prefixes permit up to four judgments, with fewer calls if the workflow stops early. One AgentVax call at a checkpoint can return risks for several horizons and choose one action. In the four-event comparison each guard acts once, at the same preselected checkpoint, under the same permissions, reviewer and deadline. We count the reviewer’s work separately.
One guard decision per workflow on Source, all-actor comparison**
<a id="benign"></a>
** False alarms, overblocking and retention at the same checkpoint**
R1a–b report the equal-count comparison. This shows that under the same and fair number of probes, agentvax can deliver better security capabilities and siginificient improvement.
Manuscript revision. We will place the matched decision counts and benign outcomes beside the Source comparison in §6.2 and Appendix E.2. Thank you for asking us to show both what was prevented and which legitimate tasks still completed.