Quantization Effects on Tool-Failure Recovery Vary Across Prompts and Evaluation Designs
cs.AI, cs.CL, cs.LG
Submitted: 2026-10-06
Updated: 2026-10-06
Code: https://github.com/ggml-org/llama.cpp
Terminology
Sources
- $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
- The case for 4-bit precision: k-bit Inference Scaling Laws
- Can Compressed LLMs Truly Act? An Empirical Evaluation of Agentic Capabilities in LLM Compression
- The Llama 3 Herd of Models
- Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMs
- Flat Score, Amplified Failures: How the Error Budget Masks Damage in Quantized LLM Agents
- Brittlebench: Quantifying LLM robustness via prompt sensitivity
- Diagnosis Before Recovery: Turning Agent Failures into Selective Self-Correction
- Qwen2.5 Technical Report
- $\tau$-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains
- When Tools Fail: Benchmarking Dynamic Replanning and Anomaly Recovery in LLM Agents
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems