Baidu's ERNIE Evolve automates research, lifting inference throughput 33.7% in three days
IgorCarron · x · 2026-11-11
Baidu's ERNIE Evolve is an automated research system aimed at recursive self-improvement: it edits code, runs experiments and uses the results to decide what to try next.
One reported result: 33.7% higher inference throughput after roughly three days of automated research, while preserving bitwise-identical tokens and log probabilities. The system uses two loops — an outer loop that picks research directions and an inner loop that changes code and tests candidates — against a fixed evaluator.
Related event: Baidu Releases Open-Source ERNIE Evolve Automated Research System(7 posts)→
More from AGI Musings
- David Patterson: Within a Year Most Will Believe SI Can Do All Jobs — davidpattersonx · 2026-11-11
- AI experts may be better at predicting RSI timelines than enterprise adoption — AndyMasley · 2026-11-11
- TaskNet: AI-Led Worker Interviews Add 79 Tasks Missing from O*NET, Raise AI Exposure for Lawyers and Devs — stanfordnlp · 2026-11-11
- Researcher: Weakly Recursive Models May Blur the Line to Self-Improvement — S_OhEigeartaigh · 2026-11-11
- Reid Hoffman teases Possible episode with OpenAI chief economist Ronnie Chatterji — reidhoffman · 2026-11-11
- BVSR: one equation to measure AI-scientist creativity, beating baselines across fields — james_y_zou · 2026-11-11