Testing Causal Models of Word Meaning in GPT-3 and -4

• URL: http://arxiv.org/abs/2305.14630v1
• Date: Wed, 24 May 2023 02:03:23 GMT
• Title: Testing Causal Models of Word Meaning in GPT-3 and -4
• Authors: Sam Musker, Ellie Pavlick
• Abstract summary: This paper evaluates the lexical representations of GPT-3 and GPT-4 through the lens of HIPE theory. We find no evidence that GPT-3 encodes the causal structure hypothesized by HIPE, but do find evidence that GPT-4 encodes such structure.
• Score: 18.654373173232205
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models (LLMs) have driven extraordinary improvements in NLP. However, it is unclear how such models represent lexical concepts-i.e., the meanings of the words they use. This paper evaluates the lexical representations of GPT-3 and GPT-4 through the lens of HIPE theory, a theory of concept representations which focuses on representations of words describing artifacts (such as "mop", "pencil", and "whistle"). The theory posits a causal graph that relates the meanings of such words to the form, use, and history of the objects to which they refer. We test LLMs using the same stimuli originally used by Chaigneau et al. (2004) to evaluate the theory in humans, and consider a variety of prompt designs. Our experiments concern judgements about causal outcomes, object function, and object naming. We find no evidence that GPT-3 encodes the causal structure hypothesized by HIPE, but do find evidence that GPT-4 encodes such structure. The results contribute to a growing body of research characterizing the representational capacity of large language models.

Related papers

• A Complexity-Based Theory of Compositionality [53.025566128892066]
  In AI, compositional representations can enable a powerful form of out-of-distribution generalization. Here, we propose a formal definition of compositionality that accounts for and extends our intuitions about compositionality. The definition is conceptually simple, quantitative, grounded in algorithmic information theory, and applicable to any representation.
  arXiv  Detail & Related papers  (2024-10-18T18:37:27Z)
• A generative framework to bridge data-driven models and scientific theories in language neuroscience [84.76462599023802]
  We present generative explanation-mediated validation, a framework for generating concise explanations of language selectivity in the brain. We show that explanatory accuracy is closely related to the predictive power and stability of the underlying statistical models.
  arXiv  Detail & Related papers  (2024-10-01T15:57:48Z)
• ExpLLM: Towards Chain of Thought for Facial Expression Recognition [61.49849866937758]
  We propose a novel method called ExpLLM to generate an accurate chain of thought (CoT) for facial expression recognition. Specifically, we have designed the CoT mechanism from three key perspectives: key observations, overall emotional interpretation, and conclusion. In experiments on the RAF-DB and AffectNet datasets, ExpLLM outperforms current state-of-the-art FER methods.
  arXiv  Detail & Related papers  (2024-09-04T15:50:16Z)
• The Geometry of Categorical and Hierarchical Concepts in Large Language Models [15.126806053878855]
  We show how to extend the formalization of the linear representation hypothesis to represent features (e.g., is_animal) as vectors. We use the formalization to prove a relationship between the hierarchical structure of concepts and the geometry of their representations. We validate these theoretical results on the Gemma and LLaMA-3 large language models, estimating representations for 900+ hierarchically related concepts using data from WordNet.
  arXiv  Detail & Related papers  (2024-06-03T16:34:01Z)
• "You Are An Expert Linguistic Annotator": Limits of LLMs as Analyzers of Abstract Meaning Representation [60.863629647985526]
  We examine the successes and limitations of the GPT-3, ChatGPT, and GPT-4 models in analysis of sentence meaning structure. We find that models can reliably reproduce the basic format of AMR, and can often capture core event, argument, and modifier structure. Overall, our findings indicate that these models out-of-the-box can capture aspects of semantic structure, but there remain key limitations in their ability to support fully accurate semantic analyses or parses.
  arXiv  Detail & Related papers  (2023-10-26T21:47:59Z)
• Can Large Language Models Explain Themselves? A Study of LLM-Generated Self-Explanations [14.685170467182369]
  Large language models (LLMs) such as ChatGPT have demonstrated superior performance on a variety of natural language processing (NLP) tasks. Since these models are instruction-tuned on human conversations to produce "helpful" responses, they can and often will produce explanations along with the response.
  arXiv  Detail & Related papers  (2023-10-17T12:34:32Z)
• Do Large GPT Models Discover Moral Dimensions in Language Representations? A Topological Study Of Sentence Embeddings [0.7416846035207727]
  We take a look at the topological structure of neuronal activity in the "brain" of Chat-GPT's foundation language model, and analyze it with respect to a metric representing the notion of fairness. We first compute a fairness metric, inspired by social literature, to identify factors that typically influence fairness assessments in humans, such as legitimacy, need, and responsibility. Our results show that sentence embeddings based on GPT-3.5 can be decomposed into two submanifolds corresponding to fair and unfair moral judgments.
  arXiv  Detail & Related papers  (2023-09-17T23:38:39Z)
• Instructed to Bias: Instruction-Tuned Language Models Exhibit Emergent Cognitive Bias [57.42417061979399]
  Recent studies show that instruction tuning (IT) and reinforcement learning from human feedback (RLHF) improve the abilities of large language models (LMs) dramatically. In this work, we investigate the effect of IT and RLHF on decision making and reasoning in LMs. Our findings highlight the presence of these biases in various models from the GPT-3, Mistral, and T5 families.
  arXiv  Detail & Related papers  (2023-08-01T01:39:25Z)
• Not wacky vs. definitely wacky: A study of scalar adverbs in pretrained language models [0.0]
  Modern pretrained language models, such as BERT, RoBERTa and GPT-3 hold the promise of performing better on logical tasks than classic static word embeddings. We investigate the extent to which BERT, RoBERTa, GPT-2 and GPT-3 exhibit general, human-like, knowledge of these common words. We find that despite capturing some aspects of logical meaning, the models fall far short of human performance.
  arXiv  Detail & Related papers  (2023-05-25T18:56:26Z)
• OPT-R: Exploring the Role of Explanations in Finetuning and Prompting for Reasoning Skills of Large Language Models [48.412284346337344]
  We conduct a thorough investigation into the reasoning capabilities of Large Language Models (LLMs) Our study entails finetuning three different sizes of Open Pretrained Transformers (OPT) We then evaluate all models on 57 out-of-domain tasks drawn from the SUPER-NATURALINSTRUCTIONS benchmark.
  arXiv  Detail & Related papers  (2023-05-19T20:58:22Z)
• Systematicity in GPT-3's Interpretation of Novel English Noun Compounds [7.039267642892591]
  We compare Levin et al.'s experimental data with GPT-3 generations, finding a high degree of similarity. We fail to find convincing evidence that GPT-3 is reasoning about more than just individual lexical items. These results highlight the importance of controlling for low-level distributional regularities when assessing whether a large language model latently encodes a deeper theory.
  arXiv  Detail & Related papers  (2022-10-18T00:25:24Z)
• Few-Shot Semantic Parsing with Language Models Trained On Code [52.23355024995237]
  We find that Codex performs better at semantic parsing than equivalent GPT-3 models. We find that unlike GPT-3, Codex performs similarly when targeting meaning representations directly, perhaps as meaning representations used in semantic parsing are structured similar to code.
  arXiv  Detail & Related papers  (2021-12-16T08:34:06Z)

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.