- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
The paper introduces what it calls "hypothesis-space inadequacy": A_ϵ (O,Q)=∅, meaning that none of the currently available hypotheses is sufficiently compatible with the observed consequences. The key idea is to do the following: (a) Score every candidate against observed consequences. (b) Test whether any candidate appears adequate. (c) Test whether the selected candidate is trustworthy. (d) If either test fails, refuse commitment and request either: revise the hypothesis set or get more evidence. That is essentially an open-set verification layer built on conformal calibration.
The paper thus is mainly a conformal-verification framework for detecting when the currently available hypothesis set may be fundamentally inadequate, and for routing between abstention, revision, and additional evidence gathering. That is a legitimate contribution, but much narrower than the paper's framing suggests.
Critical strengths
The paper separates Selection failure ("The correct explanation is represented but not chosen.") from Hypothesis-space inadequacy ("No represented explanation is task-compatible.")
Other literature addressing such tasks usually assumes a complete state space. For example, the medical diagnosis literature usually assumes the true fault belongs to the fault universe. This paper explicitly addresses when neither assumption holds. That is actually useful. The novelty is therefore not sensing, but is "calibrated detection that the current explanatory universe may itself be wrong."
The second contribution is the use of causality to integrate with information-theoretic concepts. Causality can bring key insights into such decision making processes.
Critical weaknesses
The primary issue with this paper is that it over-inflates what is claims to do, and in fact its contributions are quite narrow: mainly a conformal-verification framework for detecting when the currently available hypothesis set may be fundamentally inadequate.
Within the broader context of the defined task, the paper omits a great deal. The following comments examine what seems to be missing, relative to the other approaches to solving this task.
This paper makes a set of assumptions totally different to those made in similar active sensing contexts, viewed from a control perspective. Because the paper completely separates diagnosis, evidence gathering, decision making, this seems odd, since a control theory typically couples them. A control theorist would formulate something like π^*=argmin┬π E["cost" ] over: sensing actions diagnosis beliefs interventions future observations simultaneously. Examples of the areas taking such a perspective include POMDP diagnosis, dual control, Bayes-adaptive control, active fault diagnosis and information-state control. In all of these formulations, information gathering is valuable only because it improves future decisions.
The paper never solves such an optimization problem. Instead it proposes a cycle consisting of: verify, perhaps query, perhaps revise, perhaps commit. The query policy is argmax┬q U(q)-λc(q), but this is only a local heuristic for choosing among candidate queries, not a global optimal sensing-and-decision problem. From a control viewpoint this is extremely weak.
Why active-sensing researchers may be unimpressed: The paper repeatedly discusses acquiring additional evidence. At first glance this sounds like active sensing. However active sensing normally solves: q^=argmaxE ["future utility" ] or q^=argmax"VoI" (q) under costs. The present paper does not derive any optimal acquisition policy. Instead: query only happens after verification, queries are selected from a fixed set, the utility function is delegated to the user, no optimality theorem exists. Thus from an active-sensing perspective this is almost a post-processing wrapper around an existing diagnostic system.
I think the paper's most significant conceptual weakness concerns notions of Value of Information (VoI). The authors state that seek distinguishing evidence among unresolved candidates. But this is exactly the classical Value of Information problem. Suppose we have candidate causes h_1,…,h_k and possible measurements q_1,…,q_m. A control/decision-theoretic formulation asks: "VoI" (q)=E["decision quality after " q]-E["decision quality before " q]. The present paper never computes this quantity. Instead it introduces a generic separation utility U(q) and leaves it unspecified. That means that the following issues are not addressed: (a) no optimal evidence theory, (b) no information-theoretic characterization, (c) no regret-optimal sensing strategy. Consequently the active-sensing component is much weaker than decades-old VoI formulations.
The sensing model is surprisingly unrealistic: The paper assumes candidate predictions can be compared against observed evidence through discrepancy measures. Effectively: d(P_h,P ̂^* ) drives everything. But where does P ̂^* come from? The paper largely treats it as observable or estimable from evidence. This ignores major diagnosis issues: sensor noise, sensor bias, adversarial corruption, missing observations, partial observability. A diagnosis researcher would immediately ask: What if the evidence itself is inadequate? The paper studies only: Is the hypothesis space inadequate? Those are not the same problem. In diagnosis theory there is usually a three-way ambiguity: wrong hypothesis, inadequate hypothesis space, inadequate observations. Only (1) and (2) are represented here.
The paper is surprisingly disconnected from diagnosability theory: The paper's central object is A_ϵ (O,Q). Compatibility depends on agreement with selected consequences. In classical diagnosis theory we would ask if (a) faults can be distinguished? (b) does evidence induce an equivalence class? (c) what is the residual ambiguity? (d) what observation partition results? The paper never develops anything resembling a diagnosability partition. In fact the uncertainty set S(O) is effectively an equivalence class of indistinguishable explanations. But the authors never analyze: partition structure, identifiability conditions, observability bounds, distinguishability guarantees. This is a missed opportunity.
The action-theoretic section feels bolted on: I was particularly unconvinced by the regret decomposition R_D=R_V+R_S, where: R_V: hypothesis-space inadequacy regret, and R_S: selection regret. Mathematically this is true because the authors define R_D=J_sel-J^* and add/subtract an intermediate term. But this is not a deep result. It is essentially a bookkeeping identity. A control theorist would expect the following issues to be addressed: (a) optimal policy characterization, (b) Bellman equation, (c) regret bounds, (d) information-state evolution. None of those appear.
Questions for authors
-
Can you precisely address the VoI issues that you contribute to?
-
What is the objective that you are trying to optimize? Can this be precisely formulated?
-
How does causality improve the proposed framework, when it seems you solve exactly the classical Value of Information problem?
9 comments
Response to Reviewer KNt3 (part 2/3)
Q1 continued: implemented acquisition
Screen-first policy. The first query maximizes candidate-prediction sensitivity to Gaussian coefficient perturbations (, 400 draws) minus between-candidate spread. If both gates pass, commit. Otherwise V failure requests the screening query and S failure the candidate-separating query. Each commitment opportunity spends half the level, with separate stage banks. Query choice uses no test labels, but calibration still needs historical failure labels. Heterogeneous condition, six repetitions:
| Policy | Correct (%) | Wrong (%) | Queries |
|---|---|---|---|
| Closed-candidate greedy VoI | 5.07 | 5.07 | 1.00 |
| Fixed screening then discrimination | 39.14 | 4.95 | 2.00 |
| Screen-first | 41.28 | 3.96 | 1.80 |
| Oracle-selected fixed pair | 43.44 | 4.96 | 2.00 |
The oracle row is an evaluated fixed pair, not an LP bound. Screen-first uses a sensitivity proxy, not oracle multipliers. It loses to the fixed pair in the noisy condition (5.89% vs 12.54%). Table 35 (p. 54) has all eight conditions; the policy is specified on p. 53 lines 2847–2861.
A stronger planner acquires better and we say so. Expanded-model VoI fits a four-component product-binomial mixture for inadequate mechanisms from historical branch outcomes and solves the two-query commit/defer problem by finite-horizon dynamic programming. Heterogeneous condition, twelve repetitions:
| Policy | Correct (%) | Wrong (%) | Queries |
|---|---|---|---|
| Expanded-model VoI | 43.85 | 3.63 | 1.64 |
| Gate-directed sensing | 42.48 | 4.04 | 1.78 |
It values screening evidence because its model represents inadequate mechanisms (Table 24 p. 45).
Protection isolated inside that planner. Hold its acquisition and learned scalar fixed and change only the calibration:
| Calibration | V acceptance (%) | Correct (%) |
|---|---|---|
| Pooled failures | 12.60 | 72.37 |
| Separate V/S banks | 9.92 | 72.27 |
Twelve-repetition means, less-inadequacy condition; Table 25 (p. 46) has all conditions. Null-score calibration needs representative failure labels and null-specific exchangeability, not a correct planner posterior. Table 31 (p. 50) repeats this for six replacement workflows. Table 4B (p. 9) and Figure 8 (p. 51) give the real-telemetry trade-off.
Control characterization and regret
Regret is an identity. With the harm of action under the operative mechanism, the action candidate induces, , and , Proposition 5 (Appendix A.8 p. 18) is . The proof adds and subtracts . Nothing rests on it as a sensing-optimality theorem.
What Proposition 2 proves. Let , , the routing gate's attainable level, its false-rejection rate on adequate spaces, correct commitment under always-revise, always-query and routing, , . Then (p. 6; proof Appendix A.9 pp. 18–19) The proof couples the policies on the same incident. With the revision verdict, and . Then and , and routing differs from the matched outcome only on the disagreement event, whose probability is at most by Theorem 1. No independence of routing errors and repair success is assumed.
Response to Reviewer KNt3 (part 3/3)
Control characterization continued
The terminal problem is solved exactly. Fix a query and finite transcript . Let , , . The best randomized terminal rule solves
and by finite LP duality
(Eq. 27 p. 41). The multipliers price the two risks separately. In the archived extension they are for query and for (p. 53).
The comparator's Bellman recursion does not carry the budgets. It solves the two-query commit/defer problem by dynamic programming under its fitted model (reward , , cost ). Its state is the model posterior. The two risk budgets are not part of that state. Replacing its native terminal decision by calibrated verification changes the optimization problem, so its optimality does not transfer to the combined rule (Table 24 p. 45; Appendix E.5).
Information state. Our gate statistics are ranks in the two null banks, , with a marginal guarantee under exchangeability within each failure population. They are verification feedback, not a sufficient predictive statistic. The predictive law is fitted separately.
Finite history
Conditional on bank sizes with IID draws and an independent test incident, define , , $\Psi(n,k,r)=\Pr\lbrace \mathrm{Bin}(n,r)\le k-1\rbrace $ and . Expected correct two-gate acceptance is exactly
(Eq. 28 p. 42): passing gate means at most bank scores lie below the test score, a binomial count, independent across gates given the transcript. Proposition 6 (p. 42): with and historical incidents with null probabilities ,
At a bank needs nine failures before any commitment. RCAEval final banks average 12.9 V and 14.4 S cases; splitting them for level selection leaves under nine V cases in 65 of 80 fold banks (Appendix E.9 p. 51). This is why a better selector can lower verified yield and why we report fixed levels.
Observation noise, bias, missingness, corruption
Section 3 (p. 4 lines 167–175) now specifies how is obtained, with measurement and collection inside the frozen workflow.
Noise raises the evidence needed, computably. In the Proposition 3 construction, let each response be replaced by a fair bit with probability and put . With responses, and . The test $\psi_n=\mathbf 1\lbrace \lvert K/n-\tfrac12\rvert\le\delta/2\rbrace $ has and by Hoeffding, so suffices. At , this is 32 at and 89 at (Appendix E.1 p. 40).
Bounded score uncertainty gives a conservative gate. If, conditional on a -null test incident, all historical statuses and latent scores lie in their intervals jointly with probability at least , then the conservative rank , with counting definite and optional historical lower bounds below the test upper bound, satisfies $\Pr(\mathrm{Commit}\mid H_0^j)\le\min\lbrace 1,\alpha_j+\delta_j\rbrace $ (Appendix E.4 pp. 43–44). Pointwise intervals alone do not suffice; is not estimated from the discrepancy. Unknown bias or corruption can be observationally identical to omission, so a failed gate requests reconsideration of the workflow, not proof that the cause is absent.
Closing
The additions address what you asked: which observations can expose an omission (Proposition 3), when repair helps (Proposition 2), which decisions finite history can release (Eq. 28, Proposition 6), and how noisy or bounded evidence enters the guarantee (E.1, E.4). They complement a stronger planner; we do not claim a general optimal-control solution. If Eq. (6) and the isolation in Table 25 address the coupling concern, we would value knowing what remains so we can respond before the discussion closes. We would be grateful for your reassessment on that basis.
Response to Reviewer KNt3 (part 1/3)
Thank you for pressing on the connection between verification and decision-focused acquisition. CAVEAT's goal is to decide whether the available explanations can support the task before committing, to separate that failure from a wrong selection, and to route the matching repair under calibrated limits. The revision states this as a constrained objective over the complete workflow (Eq. 6) and solves its terminal rule exactly (Eq. 27). A sensing planner answers a different question, which evidence to acquire next, and we do not solve that sequential problem. The two are complementary. Proposition 3 gives an information limit that binds every acquisition policy on the admissible menu, optimal or not. Table 25 shows that inside the strongest planner we built, expanded-model VoI, failure-specific calibration still lowers inadequate acceptance from 12.60% to 9.92% at unchanged yield. The planner chooses the evidence. The verifier decides what it licenses. Statements are written out below. Pages refer to the clean revised PDF.
Q2 and coupling: the objective
Let ϖ be a complete workflow (proposal, sensing, revision, selection, stopping), Cϖ a task-compatible terminal selection, ΓN its verifier calibrated on N independent historical incidents, and cϖ the acquisition cost. Equation (6) (p. 5) is maxϖPr(Cϖ,ΓN=Commit)−λE[cϖ]s.t.Pr(ΓN=Commit∣H0,ϖV)≤αV,Pr(ΓN=Commit∣H0,ϖS)≤αS. V denotes inadequacy and S a wrong selection within an adequate space. The nulls are indexed by ϖ because repairs change the terminal failure populations as well as the useful decisions, so modular implementation does not imply mathematical independence. The guarantees are marginal within each failure population, not per-incident certificates. A constrained VoI formulation can express this objective. Our question is what evidence and calibration it requires when the space can be inadequate. The original separation utility is a heuristic acquisition rule, not a solution of the global problem, and the paper now says so (p. 5 lines 236–246).
Q3 and diagnosability: what causality adds
Proposition 3 (p. 7; construction and proof in Appendix E.1 pp. 40–41). Let Y=1{U≤σ(aX+bZ)} with U uniform and σ logistic. Both represented candidates have b=0 and differ only in a∈{−2,2}. Compare an adequate population (b=0) with an inadequate one (b∈{−d,d} equally likely, d>0). With total-variation discrepancy and tolerance ϵ<σ(d)−1/2 the second population is inadequate. If a and the initial information have the same joint law in both populations, every finite-horizon adaptive policy restricted to do(X=x,Z=0) has identical transcript laws in the two populations. Hence Pr(Commit∣H0V)≤αV forces Pr(correct commitment∣(H0V)c)≤αV.
In words. Interventions on X distinguish the listed candidates but cannot reveal the Z effect both omit, because under Z=0 it never enters the response. Any planner on that menu sees the same data whether or not the space is adequate. Interventions involving Z expose the omission. With m responses to do(X=0,Z=1) and K positives, K∣(H0V)c∼Bin(m,21),K∣H0V∼21Bin(m,σ(d))+21Bin(m,σ(−d)). For m=1 the laws agree. For m≥2 they differ. At m=12, d=2 their total variation is about .809. Equal single-response marginals do not imply equal joint transcript laws. This identifies a limit of the observation menu before any optimization over it. Lemma 2's zero-value condition requires a common conditional kernel, not equal marginals (p. 18).
Q1: beyond closed-candidate VoI
Across 96 configurations where the represented candidates predict the same response law, closed-candidate EVSI is zero analytically (max computed ∣⋅∣=3.33×10−16), yet an agreeing query attains the largest ideal constrained frontier in five of eight conditions, and in the heterogeneous condition querying z raises that frontier by 26.21 points over no query (Appendix E.12 p. 53). These are model checks, not achieved gains. Part 2 gives the implemented policies.