Decoding Intelligence: A Framework for Certifying Knowledge Comprehension in LLMs
- URL: http://arxiv.org/abs/2402.15929v2
- Date: Mon, 07 Oct 2024 15:01:48 GMT
- Title: Decoding Intelligence: A Framework for Certifying Knowledge Comprehension in LLMs
- Authors: Isha Chaudhary, Vedaant V. Jain, Gagandeep Singh,
- Abstract summary: We propose the first framework to certify knowledge comprehension in Large Language Models with formal probabilistic guarantees.
We design and certify novel specifications that precisely represent distributions of knowledge comprehension prompts leveraging knowledge graphs.
We find that the knowledge comprehension capability improves significantly with scaling the size of the models.
- Score: 3.6293956720749425
- License:
- Abstract: Knowledge comprehension capability is an important aspect of human intelligence. As Large Language Models (LLMs) are being envisioned as superhuman agents, it is crucial for them to be proficient at knowledge comprehension. However, existing benchmarking studies do not provide consistent, generalizable, and formal guarantees on the knowledge comprehension capabilities of LLMs. In this work, we propose the first framework to certify knowledge comprehension in LLMs with formal probabilistic guarantees. Our certificates are quantitative -- they consist of high-confidence, tight bounds on the probability that a target LLM gives the correct answer on any knowledge comprehension prompt sampled from a distribution. We design and certify novel specifications that precisely represent distributions of knowledge comprehension prompts leveraging knowledge graphs. We certify SOTA LLMs for specifications over the Wikidata5m knowledge graph. We find that the knowledge comprehension capability improves significantly with scaling the size of the models.
Related papers
- Decoding Knowledge in Large Language Models: A Framework for Categorization and Comprehension [14.039653386385519]
Large language models (LLMs) acquire, retain, and apply knowledge.
This paper introduces a novel framework, K-(CSA)2, which categorizes LLM knowledge along two dimensions: correctness and confidence.
arXiv Detail & Related papers (2025-01-02T16:34:10Z) - KaLM: Knowledge-aligned Autoregressive Language Modeling via Dual-view Knowledge Graph Contrastive Learning [74.21524111840652]
This paper proposes textbfKaLM, a textitKnowledge-aligned Language Modeling approach.
It fine-tunes autoregressive large language models to align with KG knowledge via the joint objective of explicit knowledge alignment and implicit knowledge alignment.
Notably, our method achieves a significant performance boost in evaluations of knowledge-driven tasks.
arXiv Detail & Related papers (2024-12-06T11:08:24Z) - Prompting Large Language Models with Knowledge Graphs for Question Answering Involving Long-tail Facts [50.06633829833144]
Large Language Models (LLMs) are effective in performing various NLP tasks, but struggle to handle tasks that require extensive, real-world knowledge.
We propose a benchmark that requires knowledge of long-tail facts for answering the involved questions.
Our experiments show that LLMs alone struggle with answering these questions, especially when the long-tail level is high or rich knowledge is required.
arXiv Detail & Related papers (2024-05-10T15:10:20Z) - Towards Reliable Latent Knowledge Estimation in LLMs: Zero-Prompt Many-Shot Based Factual Knowledge Extraction [15.534647327246239]
We propose to eliminate prompt engineering when probing large language models (LLMs) for factual knowledge.
Our approach, called Zero-Prompt Latent Knowledge Estimator (ZP-LKE), leverages the in-context learning ability of LLMs.
We perform a large-scale evaluation of the factual knowledge of a variety of open-source LLMs over a large set of relations and facts from the Wikidata knowledge base.
arXiv Detail & Related papers (2024-04-19T15:40:39Z) - FAC$^2$E: Better Understanding Large Language Model Capabilities by Dissociating Language and Cognition [56.76951887823882]
Large language models (LLMs) are primarily evaluated by overall performance on various text understanding and generation tasks.
We present FAC$2$E, a framework for Fine-grAined and Cognition-grounded LLMs' Capability Evaluation.
arXiv Detail & Related papers (2024-02-29T21:05:37Z) - Can Language Models Act as Knowledge Bases at Scale? [24.99538360485476]
Large language models (LLMs) have demonstrated remarkable proficiency in understanding and generating responses to complex queries.
Our research investigates whether LLMs can effectively store, recall, and reason with knowledge on a large scale comparable to latest knowledge bases (KBs) such as Wikidata.
arXiv Detail & Related papers (2024-02-22T04:20:14Z) - KoLA: Carefully Benchmarking World Knowledge of Large Language Models [87.96683299084788]
We construct a Knowledge-oriented LLM Assessment benchmark (KoLA)
We mimic human cognition to form a four-level taxonomy of knowledge-related abilities, covering $19$ tasks.
We use both Wikipedia, a corpus prevalently pre-trained by LLMs, along with continuously collected emerging corpora, to evaluate the capacity to handle unseen data and evolving knowledge.
arXiv Detail & Related papers (2023-06-15T17:20:46Z) - Do Large Language Models Know What They Don't Know? [74.65014158544011]
Large language models (LLMs) have a wealth of knowledge that allows them to excel in various Natural Language Processing (NLP) tasks.
Despite their vast knowledge, LLMs are still limited by the amount of information they can accommodate and comprehend.
This study aims to evaluate LLMs' self-knowledge by assessing their ability to identify unanswerable or unknowable questions.
arXiv Detail & Related papers (2023-05-29T15:30:13Z) - Knowledge Rumination for Pre-trained Language Models [77.55888291165462]
We propose a new paradigm dubbed Knowledge Rumination to help the pre-trained language model utilize related latent knowledge without retrieving it from the external corpus.
We apply the proposed knowledge rumination to various language models, including RoBERTa, DeBERTa, and GPT-3.
arXiv Detail & Related papers (2023-05-15T15:47:09Z) - KMIR: A Benchmark for Evaluating Knowledge Memorization, Identification
and Reasoning Abilities of Language Models [28.82149012250609]
We propose a benchmark, named Knowledge Memorization, Identification, and Reasoning test (KMIR)
KMIR covers 3 types of knowledge, including general knowledge, domain-specific knowledge, and commonsense, and provides 184,348 well-designed questions.
Preliminary experiments with various representative pre-training language models on KMIR reveal many interesting phenomenons.
arXiv Detail & Related papers (2022-02-28T03:52:57Z)
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.