tweetindex

Pasquale Minervini

@PMinervini · Edinburgh, United Kingdom · joined 02 Mar 2012

Research in ML/NLP at @EdinburghNLP (tenured faculty at @EdinburghUni), Co-Founder @Miniml_AI, @ELLISforEurope Scholar, https://t.co/5dUI3EFMmW

10 270Followers
5 073Following
11 975Posts total
37.9KViews on collected posts

Against accounts of the same size

6 posts from the last 90 days, next to the 10K–100K follower range. shown widely, but few of those viewers react.

Median views1 135this account942median for 10K–100K
Reach, %11.05%this account3.24%median for 10K–100K
Engagement, %0.97%this account1.99%median for 10K–100K
MetricThis accountMedian for 10K–100KRatio
Median views per post1 1359421.20×
Reach (views ÷ followers)11.05%3.24%3.41×
Engagement rate0.97%1.99%0.49×

Others in this range →   Compare with another account →   How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

20.1K13 Mar
1.4K1 Sep
3.4K3 Sep
920
680
10.8K4 Sep
701

Last 7 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.31%13 Mar
1.41%1 Sep
1.98%3 Sep
0.76%
0.74%
1.19%4 Sep
0.71%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.99%.

What the audience does

Likes70.2%210 in total
Reposts16.1%48 in total
Replies9.4%28 in total
Quotes2.3%7 in total
Bookmarks2.0%6 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Latest posts

hidden test sets! I've received a ton of requests for even the data used in this quick blog post (link in thread) -- I think we're overfitting on test sets https://t.co/hkKMjbI62L 701 views · 3 likes · 1 reposts · 1 replies 04 Sep 2026 · Open on X →
so: - any task/environment that you can score reliably and invest enough money (= an obscene amount) in bruteforcing its training, will be solved - every challenging and popular benchmark will meet this bar - abilities won't necessarily transfer how can we measure progress? 10.8K views · 97 likes · 8 reposts · 19 replies 04 Sep 2026 · Open on X →
llama-server -hf ggml-org/Qwen3.8-27B-GGUF:Q8_0 680 views · 5 likes · 0 reposts · 0 replies 03 Sep 2026 · Open on X →
Fazl and his group are stellar; consider applying!!! 🚀🚀🚀 920 views · 6 likes · 1 reposts · 0 replies 03 Sep 2026 · Open on X →
Hiring! Our Technical Safety and Governance (TSG) Lab at Oxford is hiring a Senior Researcher to work on interpretability, evaluations and AI safety for continually learning systems. Please apply by Sep 22 and share with anyone who might be interested. DM/email with questions! 3.4K views · 43 likes · 18 reposts · 4 replies 03 Sep 2026 · Open on X →
Whether a model is faithful (and, e.g., whether it's relying on a cue) depends heavily on environmental factors, and faithfulness/mechinterp/safety research should take that into account! 🚀 Amazing work by @aryopg et al. 1.4K views · 16 likes · 3 reposts · 0 replies 01 Sep 2026 · Open on X →
Please share it within your circles! https://t.co/mCA2NGreig 20.1K views · 40 likes · 17 reposts · 4 replies 13 Mar 2025 · Open on X →

Similar accounts