Pando: Do Interpretability Methods Work When Models Won't Explain Themselves?

• URL: http://arxiv.org/abs/2604.11061v1
• Date: Mon, 13 Apr 2026 06:42:24 GMT
• Title: Pando: Do Interpretability Methods Work When Models Won't Explain Themselves?
• Abstract summary: We introduce Pando, a model-organism benchmark that breaks this confound via an explanation axis.<n>Pando predicts held-out model decisions from $10$ labeled query-response pairs.<n>When explanations are faithful, black-box elicitation matches or exceeds all white-box methods.
• Score: 53.07826484214082
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mechanistic interpretability is often motivated for alignment auditing, where a model's verbal explanations can be absent, incomplete, or misleading. Yet many evaluations do not control whether black-box prompting alone can recover the target behavior, so apparent gains from white-box tools may reflect elicitation rather than internal signal; we call this the elicitation confounder. We introduce Pando, a model-organism benchmark that breaks this confound via an explanation axis: models are trained to produce either faithful explanations of the true rule, no explanation, or confident but unfaithful explanations of a disjoint distractor rule. Across 720 finetuned models implementing hidden decision-tree rules, agents predict held-out model decisions from $10$ labeled query-response pairs, optionally augmented with one interpretability tool output. When explanations are faithful, black-box elicitation matches or exceeds all white-box methods; when explanations are absent or misleading, gradient-based attribution improves accuracy by 3-5 percentage points, and relevance patching, RelP, gives the largest gains, while logit lens, sparse autoencoders, and circuit tracing provide no reliable benefit. Variance decomposition suggests gradients track decision computation, which fields causally drive the output, whereas other readouts are dominated by task representation, biases toward field identity and value. We release all models, code, and evaluation infrastructure.

Related papers

• Interpretability Without Tradeoffs: Disentangling Polysemanticity At Equal Predictive Performance [58.49623467782041]
  ELUDe is a method for improving the interpretability of deep neural networks.<n>It breaks latent representations into clear, inspectable sub-units that behave like interpretable features.<n>It requires no explicit training, no labels, and can be applied to pretrained models.
  arXiv  Detail & Related papers  (2026-05-29T13:39:09Z)
• Learn to Rank: Visual Attribution by Learning Importance Ranking [58.69028273772474]
  We propose a new approach to generate visual attribution maps for computer vision models.<n>Our method produces dense, pixel-level attributions in a single forward pass with optional, few-step gradient refinement.<n>Our experiments demonstrate consistent quantitative improvements and sharper, boundary-aligned explanations.
  arXiv  Detail & Related papers  (2026-04-07T12:53:22Z)
• UNBOX: Unveiling Black-box visual models with Natural-language [50.433977345786055]
  We introduce UNBOX, a framework for class-wise model dissection under fully data-free, gradient-free, and backpropagation-free constraints.<n>We evaluate UNBOX on ImageNet-1K, Waterbirds, and CelebA through semantic fidelity tests, visual-feature correlation analyses, and slice-discovery auditing.
  arXiv  Detail & Related papers  (2026-03-09T17:16:39Z)
• Reasoning Models Know When They're Right: Probing Hidden States for Self-Verification [23.190823296729732]
  We study whether reasoning models encode information about answer correctness through probing the model's hidden states.<n>The resulting probe can verify intermediate answers with high accuracy and produces highly calibrated scores.
  arXiv  Detail & Related papers  (2025-04-07T18:42:01Z)
• Predicting the Performance of Black-box LLMs through Self-Queries [60.87193950962585]
  Large language models (LLMs) are increasingly relied on in AI systems, predicting when they make mistakes is crucial.<n>In this paper, we extract features of LLMs in a black-box manner by using follow-up prompts and taking the probabilities of different responses as representations.<n>We demonstrate that training a linear model on these low-dimensional representations produces reliable predictors of model performance at the instance level.
  arXiv  Detail & Related papers  (2025-01-02T22:26:54Z)
• DISCRET: Synthesizing Faithful Explanations For Treatment Effect Estimation [21.172795461188578]
  We propose DISCRET, a self-interpretable ITE framework that synthesizes faithful, rule-based explanations for each sample. A key insight behind DISCRET is that explanations can serve dually as database queries to identify similar subgroups of samples. We provide a novel RL algorithm to efficiently synthesize these explanations from a large search space.
  arXiv  Detail & Related papers  (2024-06-02T04:01:08Z)
• Plausible Extractive Rationalization through Semi-Supervised Entailment Signal [29.67884478799914]
  We take a semi-supervised approach to optimize for the plausibility of extracted rationales. We adopt a pre-trained natural language inference (NLI) model and further fine-tune it on a small set of supervised rationales. We show that, by enforcing the alignment agreement between the explanation and answer in a question-answering task, the performance can be improved without access to ground truth labels.
  arXiv  Detail & Related papers  (2024-02-13T14:12:32Z)
• Discriminative Feature Attributions: Bridging Post Hoc Explainability and Inherent Interpretability [29.459228981179674]
  Post hoc explanations incorrectly attribute high importance to features that are unimportant or non-discriminative for the underlying task. Inherently interpretable models, on the other hand, circumvent these issues by explicitly encoding explanations into model architecture. We propose Distractor Erasure Tuning (DiET), a method that adapts black-box models to be robust to distractor erasure.
  arXiv  Detail & Related papers  (2023-07-27T17:06:02Z)
• Black-Box Anomaly Attribution [13.455748795087493]
  When a black-box machine learning model deviates from the true observation, what can be said about the reason behind that deviation? This is a fundamental and ubiquitous question that the end user in a business or industrial AI application often asks. We propose a novel likelihood-based attribution framework we call the likelihood compensation''
  arXiv  Detail & Related papers  (2023-05-29T01:42:32Z)
• Explanation Selection Using Unlabeled Data for Chain-of-Thought Prompting [80.9896041501715]
  Explanations that have not been "tuned" for a task, such as off-the-shelf explanations written by nonexperts, may lead to mediocre performance. This paper tackles the problem of how to optimize explanation-infused prompts in a blackbox fashion.
  arXiv  Detail & Related papers  (2023-02-09T18:02:34Z)
• Contrastive Explanations for Model Interpretability [77.92370750072831]
  We propose a methodology to produce contrastive explanations for classification models. Our method is based on projecting model representation to a latent space. Our findings shed light on the ability of label-contrastive explanations to provide a more accurate and finer-grained interpretability of a model's decision.
  arXiv  Detail & Related papers  (2021-03-02T00:36:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.