Agentified Assessment of Logical Reasoning Agents
Organizations: University of California, Berkeley
Abstract
We present a framework for evaluating and benchmarking logical reasoning agents when assessment itself must be reproducible, auditable, and robust to execution failures. Building on agentified assessment, we use an assessor agent to issue tasks, enforce execution budgets, parse outputs, and record structured failure types, while the agent under test only needs to expose a standardized agent-to-agent interface. As a case study, we benchmark an auto-formalization agent for first-order logic (FOL) reasoning on a solver-verified and repaired split of FOLIO. The agent translates natural language premises and conclusions into executable Z3Py programs and employs satisfiability modulo theories (SMT) solving to determine logical entailment. On the cleaned FOLIO validation set, the auto-formalization agent achieves 86.70% accuracy under the assessor protocol, outperforming a chain-of-thought baseline (73.89%).
Figures & tables
| Chain-of-Thought | Auto-formalization | |||||
|---|---|---|---|---|---|---|
| Category | Correct | Total | Acc. | Correct | Total | Acc. |
| True | 65 | 73 | 89.04% | 66 | 73 | 90.41% |
| False | 27 | 61 | 44.26% | 47 | 61 | 77.05% |
| Uncertain | 58 | 69 | 84.06% | 63 | 69 | 91.30% |
| Overall | 150 | 203 | 73.89% | 176 | 203 | 86.70% |
LLM Agent Evaluation
Large language model agent evaluation measures whether tool-using systems complete tasks reliably, safely, and efficiently.
See all 1,194 papers