Trendoday

DaisyRose-27B: 1,200-task RL took Qwen3.8 from 5.6% to 12% on Terminal-Bench 4.1

discuss

  • Points193
  • Comments63
–Rank now
#4Peak rank
2Days on list
201Score · 7d
Nov 18, 2026First seen
Rank · 7 days

On Hacker News

Points
193
Comments
63
Submitted by
de_sty
Posted
Nov 18, 2026
Site
radixark.com

Story details from Hacker News ↗ · linked article on radixark.com.

Trending history

DaisyRose-27B: 1,200-task RL took Qwen3.8 from 5.6% to 12% on Terminal-Bench 4.1 was first seen on Hacker News · Front page on Nov 18, 2026. Since then it has appeared in 72 checks on 2 different days. Its best position was #4 of 30, reached on Nov 18, 22:13 UTC. It dropped off the list after Nov 19, 13:58 UTC.

Rank history

Position on Hacker News · Front page at every check (lower is better).

#1#11#20#30 20:1300:4005:0609:3213:58 Nov 18, 20:13 UTC · #26Nov 18, 20:28 UTC · #21Nov 18, 20:43 UTC · #8Nov 18, 20:58 UTC · #6Nov 18, 21:13 UTC · #6Nov 18, 21:28 UTC · #6Nov 18, 21:43 UTC · #5Nov 18, 21:58 UTC · #5Nov 18, 22:13 UTC · #4Nov 18, 22:28 UTC · #4Nov 18, 22:43 UTC · #5Nov 18, 22:58 UTC · #5Nov 18, 23:13 UTC · #5Nov 18, 23:28 UTC · #5Nov 18, 23:43 UTC · #5Nov 18, 23:58 UTC · #4Nov 19, 00:13 UTC · #6Nov 19, 00:28 UTC · #6Nov 19, 00:43 UTC · #8Nov 19, 00:58 UTC · #9Nov 19, 01:13 UTC · #10Nov 19, 01:28 UTC · #10Nov 19, 01:43 UTC · #10Nov 19, 01:58 UTC · #11Nov 19, 02:13 UTC · #14Nov 19, 02:28 UTC · #13Nov 19, 02:43 UTC · #12Nov 19, 02:58 UTC · #12Nov 19, 03:13 UTC · #11Nov 19, 03:28 UTC · #11Nov 19, 03:43 UTC · #12Nov 19, 03:58 UTC · #12Nov 19, 04:13 UTC · #12Nov 19, 04:28 UTC · #12Nov 19, 04:43 UTC · #11Nov 19, 04:58 UTC · #11Nov 19, 05:13 UTC · #11Nov 19, 05:28 UTC · #11Nov 19, 05:43 UTC · #13Nov 19, 05:58 UTC · #9Nov 19, 06:13 UTC · #9Nov 19, 06:28 UTC · #10Nov 19, 06:44 UTC · #11Nov 19, 06:59 UTC · #11Nov 19, 07:14 UTC · #12Nov 19, 07:29 UTC · #11Nov 19, 07:44 UTC · #11Nov 19, 07:59 UTC · #11Nov 19, 08:14 UTC · #13Nov 19, 08:29 UTC · #13Nov 19, 08:44 UTC · #13Nov 19, 08:59 UTC · #13Nov 19, 09:14 UTC · #14Nov 19, 09:29 UTC · #12Nov 19, 09:44 UTC · #15Nov 19, 09:59 UTC · #14Nov 19, 10:14 UTC · #15Nov 19, 10:29 UTC · #17Nov 19, 10:44 UTC · #17Nov 19, 10:59 UTC · #17Nov 19, 11:14 UTC · #18Nov 19, 11:29 UTC · #18Nov 19, 11:44 UTC · #19Nov 19, 11:59 UTC · #18Nov 19, 12:14 UTC · #17Nov 19, 12:29 UTC · #19Nov 19, 12:44 UTC · #25Nov 19, 12:59 UTC · #24Nov 19, 13:14 UTC · #26Nov 19, 13:29 UTC · #27Nov 19, 13:44 UTC · #30Nov 19, 13:58 UTC · #30

Last 14 days

Best rank per day (UTC).

Nov 7today

Metric history

Platform metric as reported by Hacker News.

078156236 20:1300:4005:0609:3213:58 Nov 18, 20:13 UTC · 6Nov 18, 20:28 UTC · 9Nov 18, 20:43 UTC · 17Nov 18, 20:58 UTC · 30Nov 18, 21:13 UTC · 40Nov 18, 21:28 UTC · 47Nov 18, 21:43 UTC · 64Nov 18, 21:58 UTC · 71Nov 18, 22:13 UTC · 80Nov 18, 22:28 UTC · 87Nov 18, 22:43 UTC · 89Nov 18, 22:58 UTC · 94Nov 18, 23:13 UTC · 97Nov 18, 23:28 UTC · 103Nov 18, 23:43 UTC · 107Nov 18, 23:58 UTC · 112Nov 19, 00:13 UTC · 117Nov 19, 00:28 UTC · 122Nov 19, 00:43 UTC · 127Nov 19, 00:58 UTC · 129Nov 19, 01:13 UTC · 131Nov 19, 01:28 UTC · 135Nov 19, 01:43 UTC · 136Nov 19, 01:58 UTC · 138Nov 19, 02:13 UTC · 140Nov 19, 02:28 UTC · 145Nov 19, 02:43 UTC · 149Nov 19, 02:58 UTC · 150Nov 19, 03:13 UTC · 151Nov 19, 03:28 UTC · 151Nov 19, 03:43 UTC · 153Nov 19, 03:58 UTC · 156Nov 19, 04:13 UTC · 157Nov 19, 04:28 UTC · 158Nov 19, 04:43 UTC · 162Nov 19, 04:58 UTC · 163Nov 19, 05:13 UTC · 164Nov 19, 05:28 UTC · 166Nov 19, 05:43 UTC · 169Nov 19, 05:58 UTC · 174Nov 19, 06:13 UTC · 178Nov 19, 06:28 UTC · 179Nov 19, 06:44 UTC · 181Nov 19, 06:59 UTC · 184Nov 19, 07:14 UTC · 185Nov 19, 07:29 UTC · 187Nov 19, 07:44 UTC · 189Nov 19, 07:59 UTC · 191Nov 19, 08:14 UTC · 192Nov 19, 08:29 UTC · 196Nov 19, 08:44 UTC · 197Nov 19, 08:59 UTC · 200Nov 19, 09:14 UTC · 204Nov 19, 09:29 UTC · 209Nov 19, 09:44 UTC · 210Nov 19, 09:59 UTC · 212Nov 19, 10:14 UTC · 215Nov 19, 10:29 UTC · 215Nov 19, 10:44 UTC · 218Nov 19, 10:59 UTC · 219Nov 19, 11:14 UTC · 219Nov 19, 11:29 UTC · 221Nov 19, 11:44 UTC · 222Nov 19, 11:59 UTC · 223Nov 19, 12:14 UTC · 223Nov 19, 12:29 UTC · 225Nov 19, 12:44 UTC · 228Nov 19, 12:59 UTC · 231Nov 19, 13:14 UTC · 233Nov 19, 13:29 UTC · 234Nov 19, 13:44 UTC · 235Nov 19, 13:58 UTC · 236

From the discussion

A few top-level comments from Hacker News, shortened. Read the full thread on Hacker News.

Very excited to see more post-training happening on these open models. Looking forward to trying this one out once it makes its way onto OpenRouter. Qwen 27B really punches above its weight. 27B is small enough to fit on a Mac Mini / Macbook Pro which is very exciting. Starting to see more promising examples of…

The heavy reliance on eval datasets showcased here and also evidenced by the frontier lab's high payments for such data cast doubt on the claims that AI will become self-improving soon

Since some talented people with more modest means were able to do such a job, I wonder why Alibaba didn't do something like this when they trained the model. They left a lot of performance on the table. Which makes me wonder if such kind of optimizations are possible for other models like Deepseek v4.1 Flash 1017.

Questions

Is DaisyRose-27B: 1,200-task RL took Qwen3.8 from 5.6% to 12% on Terminal-Bench 4.1 trending on Hacker News right now?

Not at the moment. It was last on the list on Nov 19, 13:58 UTC, after 2 days of trending.

What is the highest rank DaisyRose-27B: 1,200-task RL took Qwen3.8 from 5.6% to 12% on Terminal-Bench 4.1 reached?

#4 on Hacker News · Front page, on Nov 18, 22:13 UTC. Trendoday checks this list every 15 minutes.

How long has DaisyRose-27B: 1,200-task RL took Qwen3.8 from 5.6% to 12% on Terminal-Bench 4.1 been trending?

It was first seen on Nov 18, 2026 and has been on the list on 2 days (72 checks) since.

How much discussion did it get on Hacker News?

193 points and 63 comments; submitted by de_sty on Nov 18, 2026.