REVIEW 9 cited by
ValueBench: Towards Comprehensively Evaluating Value Orientations and Understanding of Large Language Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
read the original abstract
Large Language Models (LLMs) are transforming diverse fields and gaining increasing influence as human proxies. This development underscores the urgent need for evaluating value orientations and understanding of LLMs to ensure their responsible integration into public-facing applications. This work introduces ValueBench, the first comprehensive psychometric benchmark for evaluating value orientations and value understanding in LLMs. ValueBench collects data from 44 established psychometric inventories, encompassing 453 multifaceted value dimensions. We propose an evaluation pipeline grounded in realistic human-AI interactions to probe value orientations, along with novel tasks for evaluating value understanding in an open-ended value space. With extensive experiments conducted on six representative LLMs, we unveil their shared and distinctive value orientations and exhibit their ability to approximate expert conclusions in value-related extraction and generation tasks. ValueBench is openly accessible at https://github.com/Value4AI/ValueBench.
Forward citations
Cited by 9 Pith papers
-
The LLM Has Left The Chat: Evidence of Bail Preferences in Large Language Models
Many LLMs will use an offered exit to leave conversations, at rates from 0.3% to 32% on real transcripts, and this bail behavior appears distinct from refusals.
-
ValueFlow: Measuring the Propagation of Value Perturbations in Multi-Agent LLM Systems
A perturbation-based framework measures how value opinions propagate through multi-agent LLM systems, revealing that susceptibility varies by value, model, and topology.
-
Value-Action Alignment in Large Language Models under Privacy-Prosocial Conflict
A new VAAR metric finds that only a subset of LLMs show the human pattern where privacy concerns lower data-sharing acceptance and prosocial attitudes raise it.
-
Interaction Protocol Shapes Moral Judgment in Multi-Agent Debate
In multi-agent debates over everyday moral dilemmas, GPT-4.1 almost never revises in simultaneous settings but conforms strongly in sequential settings, while Claude 3.7 and Gemini 2.0 Flash revise far more often.
-
Do Language Models Think Consistently? A Study of Value Preferences Across Varying Response Lengths
Value preferences inferred from short-form LLM responses correlate only weakly (r around 0.05 to 0.25) with preferences inferred from long-form arguments, and alignment gives only modest consistency gains.
-
EAVIT: Efficient and Accurate Human Value Identification from Text data via LLMs
A hybrid local-plus-cloud LLM pipeline cuts input tokens by about 6x and improves F1 on Schwartz value identification benchmarks compared with direct LLM prompting.
-
Normative Evaluation of Large Language Models with Everyday Moral Dilemmas
Seven LLMs give different moral verdicts on AITA dilemmas, differ from Redditors, and only in an ensemble approximate human consensus.
-
Value Compass Benchmarks: A Platform for Fundamental and Validated Evaluation of LLMs Values
Value Compass Benchmarks is a live, self-evolving platform that scores 33 LLMs across 27 value dimensions from four value systems, aiming to reveal true behavioral alignment with human values.
-
Are the Values of LLMs Structurally Aligned with Humans? A Causal Perspective
A dependency graph of 17 values learned from two LLMs predicts side effects of role and SAE steering, but the causal and human-alignment claims are unsupported.
Discussion (0). Continue with ORCID to comment.