Open-sourcing AstaBrief, the fast report-generation model in Asta
Summary
Ai2 is open-sourcing AstaBrief 8B, a Qwen3-8B model trained with supervised fine-tuning and direct preference optimization to write a cited scientific report in one pass from a question and retrieved excerpts.
A small open model can write the whole report.
AstaBrief turns a question and retrieved excerpts into one cited report, skipping the section-by-section Claude pipeline.
Open weights will also let institutions run AstaBrief on their own infrastructure
The recipe is SFT and DPO, not reinforcement learning.
Ai2 trained on real scientist queries, filtered synthetic reports, and preference pairs that two judges agreed on.
RL-based training can be unstable and expensive.
Citation density was the filter that worked.
Dropping low-citation synthetic examples improved grounding more than stacking several more elaborate statistics.
The strongest gains came from filtering out synthetic reports with low citation density
The speed result is about that system, not a new leaderboard.
Fast mode is about 3.5 times quicker in the Asta pipeline, and the comparisons were run against 2025 models.
Fast mode averages 51.1 seconds per report compared with 178.5 seconds for Thinking mode, about 3.5× faster.
Read the original ↗Save to your Sponge
Summarized with Sponge. Quotes are the author’s own words.