Skip to content

feat(daily): 2026-04-24 digest#18

Open
yayajjiang wants to merge 1 commit intomainfrom
digest/daily-2026-04-24
Open

feat(daily): 2026-04-24 digest#18
yayajjiang wants to merge 1 commit intomainfrom
digest/daily-2026-04-24

Conversation

@yayajjiang
Copy link
Copy Markdown
Owner

2026-04-24 Daily Digest

Two papers added to src/lib/daily.ts:

  • TEMPO: Scaling Test-time Training for Large Reasoning Models (2604.19295) — EM框架交替校准评论者与优化策略,无需标签数据即可持续提升推理——OLMO3-7B在AIME 2024从33%提升至51%,Qwen3-14B从42%提升至66%。⭐ Editor's Pick

  • Language as a Latent Variable for Reasoning Optimization (2604.21593) — 将语言选择作为潜在探索信号,仅用1.81万数学题训练,英文推理提升6.72%,常识推理提升4.9%,无需CoT标注。

https://claude.ai/code/session_01EcyX1x1Swv6d197wQKnYUw


Generated by Claude Code

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants