Mostly Reading: What Open-Book Scores Measure in Document-Level LoRA Adapters

No. 88 of 245 · arXiv · Martino M. L. Pulici, Cuong Xuan Chu, Evgeny Kharlamov, Volker Tresp · published 23 Oct 2026, 04:00 UTC · 75 min read

AI & machine learning
Paper

3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked

Researchers testing single-document LoRA adapters on Llama-3.1-8B-Instruct found that open-book scores mostly reflect a shared reading ability rather than stored document content. A random adapter recovered 82% of a specialist's open-book advantage but only 2% of its cloze advantage, which probes recall without the source text.

Read the original at arxiv.org. The judges saw only its text, with the site, the author and link domains removed; they are shown again after ranking.

Score

0.66total under rubric v2.0Contestedjudges differ by 0.47 on significance, 0.54 on reader interest

Significance0.61
Reader interest0.65
Audience breadth0.48
Practical relevance0.65
Substantiated0.90
Technical depth0.94
New information0.87
Counts against: Promotional0.03
Counts against: Overclaim0.20

Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).

Question	Judge 1	Judge 2	Judge 3	Mean	Spread	Weight	Contribution (weight times mean)
Significance	0.65	0.35	0.82	0.61	0.47 , above the contested threshold	+0.25	+0.152
Reader interest	0.60	0.40	0.94	0.65	0.54 , above the contested threshold	+0.20	+0.129
Audience breadth	0.48	0.30	0.67	0.48	0.37	+0.10	+0.048
Practical relevance	0.68	0.45	0.82	0.65	0.37	+0.15	+0.097
Substantiated	0.98	0.80	0.93	0.90	0.18	+0.10	+0.090
Technical depth	0.99	0.85	0.98	0.94	0.14	+0.10	+0.094
New information	0.91	0.70	0.99	0.87	0.29	+0.10	+0.087
Promotional	0.04	0.05	0.01	0.03	0.04	−0.15	−0.005
Overclaim	0.24	0.20	0.15	0.20	0.09	−0.15	−0.030
Sum of contributions							+0.663
Claim-check penalty							0.000
Total, as published							0.663

Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN —

Recorded for context only, checked 24 Oct 2026, 08:54 UTC. It played no part in the ranking.

Panel for this edition

Four AI agents ranked this edition: three judges scored every story blind to its site and author, and a fourth compared the top stories’ claims with the sources they cite. Summaries are written by AI.

Model versions are pinned for every edition and listed in its JSON. This one was scored under rubric v2.0; How ranking works describes the current rubric, v2.2.