Attributions All the Way Down? The Metagame of Interpretability

• URL: http://arxiv.org/abs/2605.06295v1
• Date: Thu, 07 May 2026 13:59:26 GMT
• Title: Attributions All the Way Down? The Metagame of Interpretability
• Abstract summary: We introduce the metagame, a conceptual framework for quantifying second-order interaction effects of model explanations.<n>We prove that attributions hierarchically decompose into meta-attributions, and establish these as directional extensions of existing interaction indices.
• Score: 20.626161249891993
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce the metagame, a conceptual framework for quantifying second-order interaction effects of model explanations. For any first-order attribution $φ(f)$ explaining a model $f$, we measure the directional influence of feature $j$ on the attribution of feature $i$, denoted as meta-attribution $\varphi_{j \to i}(f)$, by treating the attribution method itself as a cooperative game and computing its Shapley value. Theoretically, we prove that attributions hierarchically decompose into meta-attributions, and establish these as directional extensions of existing interaction indices. Empirically, we demonstrate that the metagame delivers insights across diverse interpretability applications: (i) quantifying token interactions in instruction-tuned language models, (ii) explaining cross-modal similarity in vision-language encoders, and (iii) interpreting text-to-image concepts in multimodal diffusion transformers.

Related papers

• PromptSplit: Revealing Prompt-Level Disagreement in Generative Models [18.957478338649114]
  Prompt-guided generative AI models have rapidly expanded across vision and language domains.<n>We propose PromptSplit, a kernel-based framework for detecting and analyzing prompt-dependent disagreement between generative models.<n>Experiments across text-to-image, text-to-text, and image-captioning settings demonstrate that PromptSplit accurately detects ground-truth behavioral differences.
  arXiv  Detail & Related papers  (2026-02-03T20:53:10Z)
• Dynamic Large Concept Models: Latent Reasoning in an Adaptive Semantic Space [56.37266873329401]
  Large Language Models (LLMs) apply uniform computation to all tokens, despite language exhibiting highly non-uniform information density.<n>We propose $textbfDynamic Large Concept Models (DLCM)$, a hierarchical language modeling framework that learns semantic boundaries from latent representations and shifts from tokens to a compressed concept space where reasoning is more efficient.
  arXiv  Detail & Related papers  (2025-12-31T04:19:33Z)
• Explaining Similarity in Vision-Language Encoders with Weighted Banzhaf Interactions [25.897711293173362]
  Language-image pre-training (LIP) enables the development of vision-language models capable of zero-shot classification, localization, multimodal retrieval, and semantic understanding.<n>Various explanation methods have been proposed to visualize the importance of input image-text pairs on the model's similarity outputs.<n>We introduce faithful interaction explanations of LIP models (FIxLIP) as a unified approach to decomposing the similarity in vision-language encoders.
  arXiv  Detail & Related papers  (2025-08-07T14:18:56Z)
• Interaction Asymmetry: A General Principle for Learning Composable Abstractions [27.749478197803256]
  We show that interaction asymmetry enables both disentanglement and compositional generalization. We propose an implementation of these criteria using a flexible Transformer-based VAE, with a novel regularizer on the attention weights of the decoder.
  arXiv  Detail & Related papers  (2024-11-12T13:33:26Z)
• Unveiling Induction Heads: Provable Training Dynamics and Feature Learning in Transformers [54.20763128054692]
  We study how a two-attention-layer transformer is trained to perform ICL on $n$-gram Markov chain data. We prove that the gradient flow with respect to a cross-entropy ICL loss converges to a limiting model.
  arXiv  Detail & Related papers  (2024-09-09T18:10:26Z)
• Explaining Text Similarity in Transformer Models [52.571158418102584]
  Recent advances in explainable AI have made it possible to mitigate limitations by leveraging improved explanations for Transformers. We use BiLRP, an extension developed for computing second-order explanations in bilinear similarity models, to investigate which feature interactions drive similarity in NLP models. Our findings contribute to a deeper understanding of different semantic similarity tasks and models, highlighting how novel explainable AI methods enable in-depth analyses and corpus-level insights.
  arXiv  Detail & Related papers  (2024-05-10T17:11:31Z)
• Understanding and Constructing Latent Modality Structures in Multi-modal Representation Learning [53.68371566336254]
  We argue that the key to better performance lies in meaningful latent modality structures instead of perfect modality alignment. Specifically, we design 1) a deep feature separation loss for intra-modality regularization; 2) a Brownian-bridge loss for inter-modality regularization; and 3) a geometric consistency loss for both intra- and inter-modality regularization.
  arXiv  Detail & Related papers  (2023-03-10T14:38:49Z)
• Disentangled Representation Learning for Text-Video Retrieval [51.861423831566626]
  Cross-modality interaction is a critical component in Text-Video Retrieval (TVR) We study the interaction paradigm in depth, where we find that its computation can be split into two terms. We propose a disentangled framework to capture a sequential and hierarchical representation.
  arXiv  Detail & Related papers  (2022-03-14T13:55:33Z)
• Cross-Modal Discrete Representation Learning [73.68393416984618]
  We present a self-supervised learning framework that learns a representation that captures finer levels of granularity across different modalities. Our framework relies on a discretized embedding space created via vector quantization that is shared across different modalities.
  arXiv  Detail & Related papers  (2021-06-10T00:23:33Z)
• Towards Unifying Feature Attribution and Counterfactual Explanations: Different Means to the Same End [17.226134854746267]
  We present a method to generate feature attribution explanations from a set of counterfactual examples. We show how counterfactual examples can be used to evaluate the goodness of an attribution-based explanation in terms of its necessity and sufficiency.
  arXiv  Detail & Related papers  (2020-11-10T05:41:43Z)

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.