Hacker GitHub

← All repos

sierra-research/tau2-bench

View on GitHub ↗ ★ 2,435 Python

τ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains

3 mentions on Hacker News by 3 people: 0 stories, 3 comments · first 2025-09-17 · latest 2026-10-04

Mentions per year 3 mentions · busiest year 2025 (2)
20082026
AllStoriesComments NewestTopOldest
  1. …the resellers have freebies, like: https://app.primeintellect.ai/dashboard/environments?ex_sort... https://github.com/sierra-research/tau2-bench
    comment by fxtentacle · 2026-10-04 · in Agents don't need memory, they need documentation
  2. …tau2-bench is flawed and 100% score is impossible, unless you add the tasks to the training set: https://github.com/sierra-research/tau2-bench/issues/89
    comment by imbusy111 · 2025-12-02 · in DeepSeek-v3.2: Pushing the frontier of open large language models [pdf]
  3. …for the feedback! In this (telecom) benchmark you can review agent policies and manuals here: 1) https://github.com/sierra-research/tau2-bench/blob/main/data... 2) https://github.com/sierra-research/tau2-bench/blob/main/data... Of course…
    comment by blndrt · 2025-09-17 · in Tau² benchmark: How a prompt rewrite boosted GPT-5-mini by 22%