tomaarsen @tomaarsen · 18 Aug 2026
Does it actually help? LightOn trained LateOn (multi-vector) & DenseOn (dense) on the same data, same 149M ModernBERT backbone, differing only in whether they pool. Multi-vector wins 9 of 13 NanoBEIR datasets: 0.6868 vs 0.6764 mean nDCG@10. Same gap on full BEIR. https://t.co/lviFm7lM82
1 259Views
28Likes
4Reposts
1Replies
0Quotes
6Bookmarks
Is that a lot?
1.98×vs this author's median637 views is typical
68Percentile for this authorof 22 recent posts
0.41×vs under 10K median3 068 views is typical
26.15%Reachviews ÷ followers
2.62%Engagement rateof viewers reacted