Palmyra x6 LLM debuts with agentic capabilities and strong benchmark performance

A new technical report details Palmyra x6, a large language model designed for agentic tasks. This model was developed using Anchored Supervised Fine-Tuning on a curated set of synthetic tool-use trajectories, employing a hybrid optimization method. Palmyra x6 demonstrates significant improvements over previous models for the Writer Agent and achieves top scores on benchmarks like BFCL Core, outperforming several recent models. AI

IMPACT This model's focus on agentic tasks and tool use could advance enterprise AI applications and automated workflows.

RANK_REASON Publication of a technical report detailing a new LLM with benchmark results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Adam
• Anchored Supervised Fine-Tuning
• BFCL Core
• mixture of experts
• muon
• Palmyra x6
• Writer Agent

• model release
• paper

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →