Agents Last Exam: Input Files for AI Agent Benchmark Tasks

Agents Last Exam (ALE) Task Input Data provides the input materials handed to AI agents at the start of each task in the ALE benchmark. The dataset is part of a family of three companion datasets published on HuggingFace by the agents-last-exam organization. It was last updated on June 5, 2026.

Use Cases

• Benchmarking AI agent performance based on structured task inputs.
• Developing agent task-handling strategies based on provided input materials.
• Analyzing the relationship between task input complexity and agent success rates.

Strengths

• Part of a structured benchmark family with three companion datasets.
• Provides browsable per-task directory layout for input files.
• Last updated on June 5, 2026.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.

Provenance

Source

agents-last-exam

Freshness

Last updated 2026-06-05 04:19:34.