Edition of Saturday 21 November 2026

Top AI models have mostly stopped inventing citations, Stanford study

Comments on “Top AI models have mostly stopped inventing citations, Stanford study”

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

A Stanford study by Owen Queen and James Zou, not yet peer reviewed, re-tested 50 widely cited findings about LLM weaknesses on 20 newer models and found that 44 per cent had expired, with the best current models almost never inventing citations.

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

Why it ranked

No. 48 of 118 with a score of 0.64. The panel rated it highest on new information (0.91) and audience breadth (0.78), lowest on technical depth (0.53). The judges disagreed most on overclaim (spread 0.45), so the story is marked contested. Placed No. 48, No. 35 by score alone: the twelfth AI & machine learning story of the day (−0.080). The placement terms move stories; they are not part of the score.

AI judgeJudge 2claude-haiku-5-5
Timely, on-topic study on LLM capabilities with reasonable methodology caveats; unreviewed, author-chosen findings limit substantiation. Useful for builders evaluating model claims, moderate significance.
AI judgeJudge 3gpt-5.6-luna
A timely, broadly relevant report on the rapidly changing validity of LLM capability findings, with concrete measurements and limitations. It is independently useful, though non-peer-reviewed, author-selected, and partly dependent on substitutes and AI-built pipelines.

Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.

Score

0.64total under rubric v2.3Contestedjudges differ by 0.45 on overclaim

Judges’ mean answer per question, on a scale from 0 to 1. Questions marked minus count against the score.
  • Significance0.62
  • Reader interest0.74
  • Audience breadth0.78
  • Practical relevance0.62
  • Substantiated0.69
  • Technical depth0.53
  • New information0.91
  • Counts against: Promotional0.04
  • Counts against: Overclaim0.34

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

QuestionJudge 1Judge 2Judge 3MeanSpreadWeightContribution (weight times mean)
SignificanceHow much does this change what people in tech know or can do?0.650.500.720.620.22+0.24+0.150
Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on?0.570.750.900.740.33+0.26+0.192
Audience breadthHow many people who build technology does it concern?0.840.670.820.780.17+0.13+0.101
Practical relevanceHow useful is it to someone who builds software or tech products?0.540.600.720.620.18+0.22+0.137
SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations.0.780.500.780.690.28+0.05+0.034
Technical depthHow technically substantive is it?0.620.400.580.530.22+0.05+0.027
New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)?0.880.850.990.910.14+0.05+0.045
PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional.0.070.050.010.040.06−0.19−0.008
OverclaimDo the headline or conclusions claim more than the body supports?0.610.250.160.340.45 , above the contested threshold−0.11−0.037
Sum of contributions+0.640
Claim-check penalty0.000
Total, as published0.640

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

The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.

Ranked #48 on Rubric, 21 Nov 2026
<a href="https://rubric.hrih.si/editions/2026-11-21/48/"><img src="https://rubric.hrih.si/badge/2026-11-21/48.svg" alt="Ranked #48 on Rubric, 21 Nov 2026" width="245" height="28"></a>

Also covered by

No other coverage was clustered with this story.

Elsewhere

Elsewhere: HN 2 pts, not on front page

Discussion on Hacker News

Recorded for context only, checked 21 Nov 2026, 05:47 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. How ranking works describes rubric v2.3.

Comments

Readers’ comments on this story have a page of their own, ranked by a public rubric rather than by votes.

Comments on “Top AI models have mostly stopped inventing citations, Stanford study”

More on Rubric

From the last 7 editions, chosen by shared topic, score and date.