Reducing the Scope of Language Models with Circuit Breakers

• URL: http://arxiv.org/abs/2410.21597v1
• Date: Mon, 28 Oct 2024 23:06:57 GMT
• Title: Reducing the Scope of Language Models with Circuit Breakers
• Authors: David Yunis, Siyu Huo, Chulaka Gunasekara, Danish Contractor,
• Abstract summary: We show that two representative language models can be poorly scoped and respond to queries they should not be addressing. We show that a recently-proposed method for general alignment, Circuit Breakers, can be adapted to scope language models to very specific tasks.
• Score: 7.464494269745494
• License:
• Abstract: Language models are now deployed in a wide variety of user-facing applications, often for specific purposes like answering questions about documentation or acting as coding assistants. As these models are intended for particular purposes, they should not be able to answer irrelevant queries like requests for poetry or questions about physics, or even worse, queries that can only be answered by humans like sensitive company policies. Instead we would like them to only answer queries corresponding to desired behavior and refuse all other requests, which we refer to as scoping. We find that, despite the use of system prompts, two representative language models can be poorly scoped and respond to queries they should not be addressing. We then conduct a comprehensive empirical evaluation of methods which could be used for scoping the behavior of language models. Among many other results, we show that a recently-proposed method for general alignment, Circuit Breakers (CB), can be adapted to scope language models to very specific tasks like sentiment analysis or summarization or even tasks with finer-grained scoping (e.g. summarizing only news articles). When compared to standard methods like fine-tuning or preference learning, CB is more robust both for out of distribution tasks, and to adversarial prompting techniques. We also show that layering SFT and CB together often results in the best of both worlds: improved performance only on relevant queries, while rejecting irrelevant ones.

Related papers

• Contextualized Evaluations: Taking the Guesswork Out of Language Model Evaluations [85.81295563405433]
  Language model users often issue queries that lack specification, where the context under which a query was issued is not explicit. We present contextualized evaluations, a protocol that synthetically constructs context surrounding an under-specified query and provides it during evaluation. We find that the presence of context can 1) alter conclusions drawn from evaluation, even flipping win rates between model pairs, 2) nudge evaluators to make fewer judgments based on surface-level criteria, like style, and 3) provide new insights about model behavior across diverse contexts.
  arXiv  Detail & Related papers  (2024-11-11T18:58:38Z)
• The Art of Saying No: Contextual Noncompliance in Language Models [123.383993700586]
  We introduce a comprehensive taxonomy of contextual noncompliance describing when and how models should not comply with user requests. Our taxonomy spans a wide range of categories including incomplete, unsupported, indeterminate, and humanizing requests. To test noncompliance capabilities of language models, we use this taxonomy to develop a new evaluation suite of 1000 noncompliance prompts.
  arXiv  Detail & Related papers  (2024-07-02T07:12:51Z)
• From RAGs to rich parameters: Probing how language models utilize external knowledge over parametric information for factual queries [6.382667978271587]
  Retrieval Augmented Generation (RAG) enriches the ability of language models to reason using external context to augment responses for a given user prompt. This approach has risen in popularity due to practical applications in various applications of language models in search, question/answering, and chat-bots. In this paper, we mechanistically examine the RAG pipeline to highlight that language models take shortcut and have a strong bias towards utilizing only the context information to answer the question, while relying minimally on their parametric memory.
  arXiv  Detail & Related papers  (2024-06-18T17:46:08Z)
• Eliciting Human Preferences with Language Models [56.68637202313052]
  Language models (LMs) can be directed to perform target tasks by using labeled examples or natural language prompts. We propose to use *LMs themselves* to guide the task specification process. We study GATE in three domains: email validation, content recommendation, and moral reasoning.
  arXiv  Detail & Related papers  (2023-10-17T21:11:21Z)
• Making Retrieval-Augmented Language Models Robust to Irrelevant Context [55.564789967211844]
  An important desideratum of RALMs, is that retrieved information helps model performance when it is relevant. Recent work has shown that retrieval augmentation can sometimes have a negative effect on performance.
  arXiv  Detail & Related papers  (2023-10-02T18:52:35Z)
• Answering Ambiguous Questions via Iterative Prompting [84.3426020642704]
  In open-domain question answering, due to the ambiguity of questions, multiple plausible answers may exist. One approach is to directly predict all valid answers, but this can struggle with balancing relevance and diversity. We present AmbigPrompt to address the imperfections of existing approaches to answering ambiguous questions.
  arXiv  Detail & Related papers  (2023-07-08T04:32:17Z)
• Regularized Contrastive Learning of Semantic Search [0.0]
  Transformer-based models are widely used as retrieval models due to their excellent ability to learn semantic representations. We propose a new regularization method: Regularized Contrastive Learning. It augments several different semantic representations for every sentence, then take them into the contrastive objective as regulators.
  arXiv  Detail & Related papers  (2022-09-27T08:25:19Z)
• Query Resolution for Conversational Search with Limited Supervision [63.131221660019776]
  We propose QuReTeC (Query Resolution by Term Classification), a neural query resolution model based on bidirectional transformers. We show that QuReTeC outperforms state-of-the-art models, and furthermore, that our distant supervision method can be used to substantially reduce the amount of human-curated data required to train QuReTeC.
  arXiv  Detail & Related papers  (2020-05-24T11:37:22Z)

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.