Pith. sign in

REVIEW 6 cited by

How predictable is language model benchmark performance?

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2401.04757 v1 pith:L62MDSBQ submitted 2024-01-09 cs.LG cs.AI

classification cs.LGcs.AI
keywords performancecomputeacrossaveragebig-benchindividualmagnitudemodel
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We investigate large language model performance across five orders of magnitude of compute scaling in eleven recent model architectures. We show that average benchmark performance, aggregating over many individual tasks and evaluations as in the commonly-used BIG-Bench dataset, is decently predictable as a function of training compute scale. Specifically, when extrapolating BIG-Bench Hard performance across one order of magnitude in compute, we observe average absolute errors of 6 percentage points (pp). By contrast, extrapolation for individual BIG-Bench tasks across an order of magnitude in compute yields higher average errors of 18pp. Nonetheless, individual task performance remains significantly more predictable than chance. Overall, our work suggests compute scaling provides a promising basis to forecast AI capabilities in diverse benchmarks, though predicting performance in specific tasks poses challenges.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. EdgeBench: Unveiling Scaling Laws of Learning from Real-World Environments

    cs.CL 2026-07 conditional novelty 7.5 of 10

    Across ~38,000 hours on 134 ultra-long real-world tasks, aggregate agent performance follows a log-sigmoid of interaction time, and measured learning speed doubles about every three months.

  2. Inference Compute-Optimal Video Vision Language Models

    cs.CV 2025-05 conditional novelty 7.0 of 10

    For video vision-language models under a fixed inference compute budget, the optimal setup scales language model size, frame count, and tokens per frame jointly, and shifts toward smaller LMs with more frames and toke...

  3. A Theory of Inference Compute Scaling: Reasoning through Directed Stochastic Skill Search

    cs.LG 2025-06 conditional novelty 6.0 of 10

    A skill-graph random-walk model gives closed-form accuracy-versus-compute formulas for four reasoning strategies and connects them to training scaling.

  4. How Benchmark Prediction from Fewer Data Misses the Mark

    cs.LG 2025-06 conditional novelty 6.0 of 10

    Benchmark prediction methods mostly work by interpolation among similar models and fail on better, unfamiliar models, where random sampling with an AIPW-style correction is the only consistent improvement.

  5. Meek Models Shall Inherit the Earth

    cs.AI 2025-07 conditional novelty 5.0 of 10

    Under fixed-distribution neural scaling laws, the capability gap between state-of-the-art and low-compute AI models shrinks over time toward zero.

  6. Capability Salience Vector: Fine-grained Alignment of Loss and Capabilities for Downstream Task Scaling Law

    cs.CL 2025-06 conditional novelty 5.0 of 10

    A fitted token-level loss weighting, optimized against known model accuracies, predicts held-out downstream task performance more accurately than mean validation loss on five of six benchmarks.

Pith tools