tweetindex

Gabriel P. Andrade @gab_p_andrade · 08 Jul 2026

9/ As always, there's no free lunch here: LLM errors aren't uniform. Their biases produce concentrated, learnable error patterns that an ex-ante attacker can chase. These error profiles become an audit surface (open weights, bias bounties, etc.)
92Views
3Likes
0Reposts
1Replies
0Quotes
1Bookmarks

Is that a lot?

0.63×vs this author's median145 views is typical
0Percentile for this authorof 11 recent posts
0.02×vs under 10K median4 686 views is typical
13.18%Reachviews ÷ followers
4.35%Engagement rateof viewers reacted

Compare with the benchmark table →

Open on X →