REVIEW 10 cited by
MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback
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
read the original abstract
To solve complex tasks, large language models (LLMs) often require multiple rounds of interactions with the user, sometimes assisted by external tools. However, current evaluation protocols often emphasize benchmark performance with single-turn exchanges, neglecting the nuanced interactions among the user, LLMs, and external tools, while also underestimating the importance of natural language feedback from users. These oversights contribute to discrepancies between research benchmark evaluations and real-world use cases. We introduce MINT, a benchmark that evaluates LLMs' ability to solve tasks with multi-turn interactions by (1) using tools and (2) leveraging natural language feedback. To ensure reproducibility, we provide an evaluation framework where LLMs can access tools by executing Python code and receive users' natural language feedback simulated by GPT-4. We repurpose a diverse set of established evaluation datasets focusing on reasoning, coding, and decision-making and carefully curate them into a compact subset for efficient evaluation. Our analysis of 20 open- and closed-source LLMs offers intriguing findings. (a) LLMs generally benefit from tools and language feedback, with performance gains (absolute, same below) of 1-8% for each turn of tool use and 2-17% with natural language feedback. (b) Better single-turn performance does not guarantee better multi-turn performance. (c) Surprisingly, on the LLMs evaluated, supervised instruction-finetuning (SIFT) and reinforcement learning from human feedback (RLHF) generally hurt multi-turn capabilities. We expect MINT can help measure progress and incentivize research in improving LLMs' capabilities in multi-turn interactions, especially for open-source communities where multi-turn human evaluation can be less accessible compared to commercial LLMs with a larger user base.
Forward citations
Cited by 10 Pith papers
-
DataClawEval: A Benchmark for Data Engineering Agents in Real Industrial Harness
A production-grounded, sandbox-graded benchmark of 100 multi-engine data-engineering tasks shows frontier agents top out at 74.9 with no cross-engine winner.
-
Lost in Execution: On the Multilingual Robustness of Tool Calling in Large Language Models
A new benchmark shows multilingual tool-calling errors in LLMs are mostly parameter-language mismatches at the execution boundary, not failures of intent understanding.
-
Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Live
TTL-based KV-cache pinning that uses predicted tool-call durations and queueing-delay costs cuts agent job completion time by up to 8x.
-
A Simple "Try Again" Can Elicit Multi-Turn LLM Reasoning
Multi-turn RL with only unary 'try again' feedback improves multi-turn reasoning accuracy by up to 14% while preserving single-turn performance.
-
DrafterBench: Benchmarking Large Language Models for Tasks Automation in Civil Engineering
DrafterBench is a new benchmark of 1,920 PDF drawing-revision tasks; on it, the best model (OpenAI o1) averages about 80/100, and all tested models fail hard on incomplete instructions and plan execution.
-
Automating Financial Statement Audits with Large Language Models
A five-stage benchmark shows LLMs detect financial statement errors well but fail at explanation, standards citation, and statement revision.
-
ChemGraph: An Agentic Framework for Computational Chemistry Workflows
A new LLM-driven framework, ChemGraph, automates molecular simulation workflows and shows that multi-agent task decomposition improves smaller models' accuracy on complex thermochemistry benchmarks.
-
A Call for Collaborative Intelligence: Why Human-Agent Systems Should Precede AI Autonomy
A position paper arguing that LLM-based human-agent systems, not fully autonomous agents, should be the immediate goal for AI development.
-
Evolutionary Perspectives on the Evaluation of LLM-Based AI Agents: A Comprehensive Survey
A survey that classifies AI agent evaluation benchmarks along environment and capability axes, and proposes five traits that distinguish agents from chatbots.
-
LLM-Based Agentic Systems for Software Engineering: Challenges and Opportunities
A survey of LLM-based multi-agent systems across the software development life cycle, plus a research agenda for orchestration, human coordination, cost, and data.
Discussion (0). Sign in to comment.