Low-Bit N-Gram Embeddings for Language Models
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
The paper proposes storing n-gram embedding tables, which some open-weight language models use to boost performance, at low bit precision to cut their memory footprint.
Score
0.69total under rubric v2.0
- Significance0.60
- Reader interest0.76
- Audience breadth0.54
- Practical relevance0.75
- Substantiated0.89
- Technical depth0.93
- New information0.90
- Counts against: Promotional0.05
- Counts against: Overclaim0.25
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 | 0.62 | 0.45 | 0.72 | 0.60 | 0.27 | +0.25 | +0.149 |
| Reader interest | 0.67 | 0.70 | 0.90 | 0.76 | 0.23 | +0.20 | +0.151 |
| Audience breadth | 0.53 | 0.50 | 0.58 | 0.54 | 0.08 | +0.10 | +0.054 |
| Practical relevance | 0.76 | 0.70 | 0.78 | 0.75 | 0.08 | +0.15 | +0.112 |
| Substantiated | 0.97 | 0.80 | 0.90 | 0.89 | 0.17 | +0.10 | +0.089 |
| Technical depth | 0.98 | 0.85 | 0.96 | 0.93 | 0.13 | +0.10 | +0.093 |
| New information | 0.90 | 0.80 | 0.99 | 0.90 | 0.19 | +0.10 | +0.090 |
| Promotional | 0.05 | 0.05 | 0.04 | 0.05 | 0.01 | −0.15 | −0.007 |
| Overclaim | 0.32 | 0.30 | 0.12 | 0.25 | 0.20 | −0.15 | −0.037 |
| Sum of contributions | +0.693 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.694 | ||||||
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, 08:53 UTC. It played no part in the ranking.