On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

count % of corpus

5 results for “compute-allocation”

Strategic Scaling of Test-Time Compute: A Bandit Learning Approachcs.AI

Adaptive compute allocation improves model performance by prioritizing harder queries.

problem Inefficiency in allocating test-time compute uniformly across all queries.

method Formulated as a bandit learning problem, proposed adaptive algorithms that estimate query difficulty and allocate compute accordingly.

result Achieved up to 15.29% relative performance improvement on various benchmarks.

test-time compute adaptive allocation bandit learning query difficulty language models

2025-06-15abs ↗pdf ↗

Learning Shrinks the Hard Tail: Training-Dependent Inference Scaling in a Solvable Linear Modelcs.LG

Learning shrinks hard tail, improving inference performance.

problem Improving inference performance in neural networks.

method Latent Instance Difficulty (LID) model analyzing fine-tuning of neural networks.

result Training-dependent inference scaling, with

β_ ext{eff}

growing with sample size before saturating.

neural scaling fine-tuning inference hard tail compute-allocation

2026-01-07abs ↗pdf ↗

BUOCA: Budget-Optimized Crowd Worker Allocationcs.HC

Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m…

2019-01-11abs ↗pdf ↗

How to Train Your LLM Web Agent: A Statistical Diagnosiscs.AI

Study optimizes compute usage for LLM web agents, improving performance.

problem High compute costs and narrow focus on single-step tasks limit LLM web agents.

method Two-stage pipeline: SFT followed by on-policy RL, with hyperparameter optimization.

result Combining SFT and on-policy RL requires 55% less compute to match peak SFT performance.

LLM web agents compute optimization reinforcement learning hyperparameter tuning work arena

2025-07-05abs ↗pdf ↗

Intrinsic Selection and Particle Resampling for Inference-Time Scaling Beyond Domain Verifiabilitycs.LG

Inference-Time Scaling can be extended to domains prone to systematic failure using intrinsic statistics.

problem Scaling inference time in domains prone to systematic failure

method Intrinsic Selection (iS), Intrinsic Particle Filtering (iPF), and Particle Distillation (dPF)

result Intrinsic Selection improves engineering design selection by 20% and pass@1 by 6.1 points on average.

Inference-Time Scaling Intrinsic Selection Intrinsic Particle Filtering Particle Distillation Engineering Design

2026-06-07abs ↗pdf ↗