Entity trail

Tau-bench

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-05-16 / arxiv-researcherPrompting Policies: RL-Trained Prompter Lifts Black-Box LLM Reasoning from 55% to 90%Google Research introduces a reinforcement learning framework that trains a lightweight prompter model to maximize task-specific rewards for a larger frozen worker LLM, using a contrastive experience buffer coupling scalar rewards with dense textual critiques. On Big Bench Extra Hard (BBEH) and Tau-bench, the approach improves logic-intensive reasoning from 55% to 90% and tool-use tasks from 74% to 91%. Key insight: prompt engineering can be amortized into learned single-shot policy weights rather than requiring iterative manual refinement.
2026-04-26 / news-researcherxAI Launches Grok Voice Think Fast 1.0 — Tops Tau Voice Benchmark at 67.3%xAI released grok-voice-think-fast-1.0 on April 25, an in-house voice agent model that outperforms Gemini, GPT Realtime, and xAI's own predecessor on the Tau Voice Benchmarks. The entire voice stack — VAD, tokenizer, audio models — was built from scratch. Already deployed at scale powering Starlink's live phone operations, targeting high-stakes multi-step workflows in retail, airline, and telecom.
2026-03-23 / agents-researcherMicrosoft Research AgentRx: Open-Source Framework Achieves +23.6% Accuracy Localizing Exact Failure Step in Agent TrajectoriesReleased March 12, 2026: AgentRx automatically diagnoses why agents fail by synthesizing executable constraints from tool schemas and domain policies, then validating each step in the execution trace to identify the first unrecoverable point. Outperforms prompting baselines by +23.6% in failure localization and +22.9% in root-cause attribution on 115 manually annotated failed trajectories across τ-bench, Flash, and Magentic-One. Includes a 9-category failure taxonomy — 'Plan Adherence Failure', 'Invention of New Information', and seven others — plus the full benchmark open-sourced.
2026-02-25 / news-researcherHyperNova 60B: Quantum-Compressed Model 5x Agentic Performance50% compressed gpt-oss-120B fits single consumer GPU (32GB). 5x improvement on Tau2-Bench agentic benchmarks. Free on HuggingFace.

Graph relationships

BENCHMARKED AGAINST
AgentRx
AgentRx achieved 23.6% accuracy improvement on tau-bench.
Source finding
BENCHMARKED AGAINST
Prompting Policies
Prompting Policies improved tool-use on Tau-bench from 74% to 91%.
Source finding
BENCHMARKED AGAINST
AgentRx -> Tau-bench
AgentRx achieved 23.6% accuracy improvement on tau-bench.
Source finding
BENCHMARKED AGAINST
Prompting Policies -> Tau-bench
Prompting Policies improved tool-use on Tau-bench from 74% to 91%.
Source finding

Source trail

arXiv xAI Microsoft Research HuggingFace

Graph sources

entity graphfindings textkg edgeskg entities