Write your own LLM Benchmarks
Comments on “Write your own LLM Benchmarks”
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
A blog post argues that teams should build their own LLM benchmarks instead of relying on public scores, which measure other people's problems and often bundle capabilities users don't need.
Why it ranked
No. 126 of 155 with a score of 0.57. The panel rated it highest on new information (0.88) and practical relevance (0.80), lowest on substantiated (0.36).
- AI judgeJudge 2claude-haiku-5-5
- Practical guide to building custom LLM benchmarks with deterministic checks and harness design; useful to builders but lightly evidenced and somewhat promotional in tone, not a landmark.}
- AI judgeJudge 3gpt-5.6-luna
- A practical tutorial on designing deterministic, reusable LLM benchmarks and harnesses, with useful architecture and cost guidance. It lacks implementation detail, empirical validation, and some examples appear illustrative rather than well-supported.
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
0.57total under rubric v2.2
- Significance0.43
- Reader interest0.64
- Audience breadth0.60
- Practical relevance0.80
- Substantiated0.36
- Technical depth0.58
- New information0.88
- Counts against: Promotional0.12
- Counts against: Overclaim0.24
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 | 0.52 | 0.20 | 0.58 | 0.43 | 0.38 | +0.24 | +0.104 |
| Reader interest | 0.61 | 0.50 | 0.82 | 0.64 | 0.32 | +0.26 | +0.167 |
| Audience breadth | 0.64 | 0.50 | 0.67 | 0.60 | 0.17 | +0.13 | +0.078 |
| Practical relevance | 0.94 | 0.60 | 0.86 | 0.80 | 0.34 | +0.22 | +0.176 |
| Substantiated | 0.37 | 0.25 | 0.47 | 0.36 | 0.22 | +0.05 | +0.018 |
| Technical depth | 0.78 | 0.40 | 0.57 | 0.58 | 0.38 | +0.05 | +0.029 |
| New information | 0.89 | 0.90 | 0.84 | 0.88 | 0.06 | +0.05 | +0.044 |
| Promotional | 0.09 | 0.20 | 0.08 | 0.12 | 0.12 | −0.19 | −0.023 |
| Overclaim | 0.21 | 0.30 | 0.22 | 0.24 | 0.09 | −0.11 | −0.027 |
| Sum of contributions | +0.567 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.567 | ||||||
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.
<a href="https://rubric.hrih.si/editions/2026-10-28/126/"><img src="https://rubric.hrih.si/badge/2026-10-28/126.svg" alt="Ranked #126 on Rubric, 28 Oct 2026" width="252" height="28"></a>Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 2 pts, not on front page
Recorded for context only, checked 28 Oct 2026, 06:47 UTC. It played no part in the ranking.