Contextualized Evaluations: Taking the Guesswork Out of Language Model Evaluations

• URL: http://arxiv.org/abs/2411.07237v1
• Date: Mon, 11 Nov 2024 18:58:38 GMT
• Title: Contextualized Evaluations: Taking the Guesswork Out of Language Model Evaluations
• Authors: Chaitanya Malaviya, Joseph Chee Chang, Dan Roth, Mohit Iyyer, Mark Yatskar, Kyle Lo,
• Abstract summary: 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.
• Score: 85.81295563405433
• License:
• Abstract: Language model users often issue queries that lack specification, where the context under which a query was issued -- such as the user's identity, the query's intent, and the criteria for a response to be useful -- is not explicit. For instance, a good response to a subjective query like "What book should I read next?" would depend on the user's preferences, and a good response to an open-ended query like "How do antibiotics work against bacteria?" would depend on the user's expertise. This makes evaluation of responses to such queries an ill-posed task, as evaluators may make arbitrary judgments about the response quality. To remedy this, we present contextualized evaluations, a protocol that synthetically constructs context surrounding an underspecified 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. Specifically, our procedure uncovers an implicit bias towards WEIRD contexts in models' "default" responses and we find that models are not equally sensitive to following different contexts, even when they are provided in prompts.

Related papers

• Reducing the Scope of Language Models with Circuit Breakers [7.464494269745494]
  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.
  arXiv  Detail & Related papers  (2024-10-28T23:06:57Z)
• QUDSELECT: Selective Decoding for Questions Under Discussion Parsing [90.92351108691014]
  Question Under Discussion (QUD) is a discourse framework that uses implicit questions to reveal discourse relationships between sentences. We introduce QUDSELECT, a joint-training framework that selectively decodes the QUD dependency structures considering the QUD criteria. Our method outperforms the state-of-the-art baseline models by 9% in human evaluation and 4% in automatic evaluation.
  arXiv  Detail & Related papers  (2024-08-02T06:46:08Z)
• "My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models [40.867655189493924]
  Open-ended nature of language generation makes evaluation of large language models (LLMs) challenging. One common evaluation approach uses multiple-choice questions (MCQ) to limit the response space. We evaluate how aligned first-token evaluation is with the text output along several dimensions.
  arXiv  Detail & Related papers  (2024-02-22T12:47:33Z)
• RLVF: Learning from Verbal Feedback without Overgeneralization [94.19501420241188]
  We study the problem of incorporating verbal feedback without such overgeneralization. We develop a new method Contextualized Critiques with Constrained Preference Optimization (C3PO) Our approach effectively applies verbal feedback to relevant scenarios while preserving existing behaviors for other contexts.
  arXiv  Detail & Related papers  (2024-02-16T18:50:24Z)
• What if you said that differently?: How Explanation Formats Affect Human Feedback Efficacy and User Perception [53.4840989321394]
  We analyze the effect of rationales generated by QA models to support their answers. We present users with incorrect answers and corresponding rationales in various formats. We measure the effectiveness of this feedback in patching these rationales through in-context learning.
  arXiv  Detail & Related papers  (2023-11-16T04:26:32Z)
• EvalLM: Interactive Evaluation of Large Language Model Prompts on User-Defined Criteria [43.944632774725484]
  We present EvalLM, an interactive system for iteratively refining prompts by evaluating multiple outputs on user-defined criteria. By describing criteria in natural language, users can employ the system's LLM-based evaluator to get an overview of where prompts excel or fail. A comparative study showed that EvalLM, when compared to manual evaluation, helped participants compose more diverse criteria, examine twice as many outputs, and reach satisfactory prompts with 59% fewer revisions.
  arXiv  Detail & Related papers  (2023-09-24T13:19:38Z)
• 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)
• Searching for Better Database Queries in the Outputs of Semantic Parsers [16.221439565760058]
  In this paper, we consider the case when, at the test time, the system has access to an external criterion that evaluates the generated queries. The criterion can vary from checking that a query executes without errors to verifying the query on a set of tests. We apply our approach to the state-of-the-art semantics and report that it allows us to find many queries passing all the tests on different datasets.
  arXiv  Detail & Related papers  (2022-10-13T17:20:45Z)
• QRelScore: Better Evaluating Generated Questions with Deeper Understanding of Context-aware Relevance [54.48031346496593]
  We propose $textbfQRelScore$, a context-aware evaluation metric for $underlinetextbfRel$evance evaluation metric. Based on off-the-shelf language models such as BERT and GPT2, QRelScore employs both word-level hierarchical matching and sentence-level prompt-based generation. Compared with existing metrics, our experiments demonstrate that QRelScore is able to achieve a higher correlation with human judgments while being much more robust to adversarial samples.
  arXiv  Detail & Related papers  (2022-04-29T07:39:53Z)
• Speaker Sensitive Response Evaluation Model [17.381658875470638]
  We propose an automatic evaluation model based on the similarity of the generated response with the conversational context. We learn the model parameters from an unlabeled conversation corpus. We show that our model can be applied to movie dialogues without any additional training.
  arXiv  Detail & Related papers  (2020-06-12T08:59:10Z)

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.