Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Models

• URL: http://arxiv.org/abs/2401.06102v4
• Date: Thu, 6 Jun 2024 22:59:58 GMT
• Title: Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Models
• Authors: Asma Ghandeharioun, Avi Caciularu, Adam Pearce, Lucas Dixon, Mor Geva,
• Abstract summary: We introduce a framework called Patchscopes and show how it can be used to answer a wide range of questions about an LLM's computation. We show that many prior interpretability methods based on projecting representations into the vocabulary space and intervening on the LLM can be viewed as instances of this framework. Beyond unifying prior inspection techniques, Patchscopes also opens up new possibilities such as using a more capable model to explain the representations of a smaller model, and multihop reasoning error correction.
• Score: 24.817659341654654
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Understanding the internal representations of large language models (LLMs) can help explain models' behavior and verify their alignment with human values. Given the capabilities of LLMs in generating human-understandable text, we propose leveraging the model itself to explain its internal representations in natural language. We introduce a framework called Patchscopes and show how it can be used to answer a wide range of questions about an LLM's computation. We show that many prior interpretability methods based on projecting representations into the vocabulary space and intervening on the LLM computation can be viewed as instances of this framework. Moreover, several of their shortcomings such as failure in inspecting early layers or lack of expressivity can be mitigated by Patchscopes. Beyond unifying prior inspection techniques, Patchscopes also opens up new possibilities such as using a more capable model to explain the representations of a smaller model, and multihop reasoning error correction.

Related papers

• LINGOLY-TOO: Disentangling Memorisation from Reasoning with Linguistic Templatisation and Orthographic Obfuscation [1.2576388595811496]
  We introduce a framework for producing linguistic reasoning problems that reduces the effect of memorisation in model performance estimates. We apply this framework to develop LINGOLY-TOO, a challenging benchmark for linguistic reasoning.
  arXiv  Detail & Related papers  (2025-03-04T19:57:47Z)
• Superscopes: Amplifying Internal Feature Representations for Language Model Interpretation [0.0]
  We introduce Superscopes, a technique that amplifies features in models into new contexts. Superscopes enables the interpretation of internal representations that previous methods failed to explain-all without requiring additional training. This approach provides new insights into how LLMs build context and represent complex concepts, further advancing mechanistic interpretability.
  arXiv  Detail & Related papers  (2025-03-03T21:58:12Z)
• 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. 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. 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)
• LatentQA: Teaching LLMs to Decode Activations Into Natural Language [72.87064562349742]
  We introduce LatentQA, the task of answering open-ended questions about model activations in natural language. We propose Latent Interpretation Tuning (LIT), which finetunes a decoder LLM on a dataset of activations and associated question-answer pairs. Our decoder also specifies a differentiable loss that we use to control models, such as debiasing models on stereotyped sentences and controlling the sentiment of generations.
  arXiv  Detail & Related papers  (2024-12-11T18:59:33Z)
• 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)
• On the Tip of the Tongue: Analyzing Conceptual Representation in Large Language Models with Reverse-Dictionary Probe [36.65834065044746]
  We use in-context learning to guide the models to generate the term for an object concept implied in a linguistic description. Experiments suggest that conceptual inference ability as probed by the reverse-dictionary task predicts model's general reasoning performance.
  arXiv  Detail & Related papers  (2024-02-22T09:45:26Z)
• Explainability for Large Language Models: A Survey [59.67574757137078]
  Large language models (LLMs) have demonstrated impressive capabilities in natural language processing. This paper introduces a taxonomy of explainability techniques and provides a structured overview of methods for explaining Transformer-based language models.
  arXiv  Detail & Related papers  (2023-09-02T22:14:26Z)
• Evaluating and Explaining Large Language Models for Code Using Syntactic Structures [74.93762031957883]
  This paper introduces ASTxplainer, an explainability method specific to Large Language Models for code. At its core, ASTxplainer provides an automated method for aligning token predictions with AST nodes. We perform an empirical evaluation on 12 popular LLMs for code using a curated dataset of the most popular GitHub projects.
  arXiv  Detail & Related papers  (2023-08-07T18:50:57Z)
• Explaining Emergent In-Context Learning as Kernel Regression [61.57151500616111]
  Large language models (LLMs) have initiated a paradigm shift in transfer learning. In this paper, we investigate the reason why a transformer-based language model can accomplish in-context learning after pre-training. We find that during ICL, the attention and hidden features in LLMs match the behaviors of a kernel regression.
  arXiv  Detail & Related papers  (2023-05-22T06:45:02Z)
• Competence-Based Analysis of Language Models [21.43498764977656]
  CALM (Competence-based Analysis of Language Models) is designed to investigate LLM competence in the context of specific tasks. We develop a new approach for performing causal probing interventions using gradient-based adversarial attacks. We carry out a case study of CALM using these interventions to analyze and compare LLM competence across a variety of lexical inference tasks.
  arXiv  Detail & Related papers  (2023-03-01T08:53:36Z)
• Shapley Head Pruning: Identifying and Removing Interference in Multilingual Transformers [54.4919139401528]
  We show that it is possible to reduce interference by identifying and pruning language-specific parameters. We show that removing identified attention heads from a fixed model improves performance for a target language on both sentence classification and structural prediction.
  arXiv  Detail & Related papers  (2022-10-11T18:11:37Z)
• Interpreting Language Models with Contrastive Explanations [99.7035899290924]
  Language models must consider various features to predict a token, such as its part of speech, number, tense, or semantics. Existing explanation methods conflate evidence for all these features into a single explanation, which is less interpretable for human understanding. We show that contrastive explanations are quantifiably better than non-contrastive explanations in verifying major grammatical phenomena.
  arXiv  Detail & Related papers  (2022-02-21T18:32:24Z)
• Foundations of Symbolic Languages for Model Interpretability [2.3361634876233817]
  We study the computational complexity of FOIL queries over two classes of ML models often deemed to be easily interpretable. We present a prototype implementation of FOIL wrapped in a high-level declarative language.
  arXiv  Detail & Related papers  (2021-10-05T21:56:52Z)
• General Pitfalls of Model-Agnostic Interpretation Methods for Machine Learning Models [1.025459377812322]
  We highlight many general pitfalls of machine learning model interpretation, such as using interpretation techniques in the wrong context. We focus on pitfalls for global methods that describe the average model behavior, but many pitfalls also apply to local methods that explain individual predictions.
  arXiv  Detail & Related papers  (2020-07-08T14:02:56Z)

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.