arXiv Machine Learning August 24, 2026 By Gihyeon Sim

When to Ponder: Adaptive Compute Allocation for Code Generation via Test-Time Training

Read the original on arXiv Machine Learning →

The paper introduces PonderTTT, a gating strategy that uses the self‑supervised reconstruction loss from a Test‑Time Training (TTT) layer to decide when to apply TTT updates during inference. The gating mechanism is training‑free, relying only on a single scalar threshold that is calibrated on unlabeled data and updated via exponential moving average to keep the desired update rate. Experiments on GPT‑2 models (124M–1.5B) for code language modeling show that this approach achieves 82–89% oracle recovery and outperforms random skip baselines, reducing loss on out‑of‑distribution languages by up to 16%.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

llms

arXiv Machine Learning

Jul 10

Selective Left-Shift: Turning Test-Time Compute and Difficulty-based Curation into Training Data for Low-Resource Code Generation

arXiv:2607. 07748v1 Announce Type: new Abstract: Large Language Models achieve strong code generation for high resource languages like Python and Java but suffer sharp performance drops on Low-Resource Programming Languages~(LRPLs) such as Julia.

By Didula Samaraweera, Anjana Supun, Srinath Perera

llms reinforcement-learning fine-tuning benchmarks

arXiv Machine Learning

4d ago

Performance, Efficiency and Collapse -- Advantages and Challenges in Offline Post-training of Code LLMs

The paper investigates whether reinforcement‑learning post‑training of code‑generating large language models can be done entirely offline using existing datasets, avoiding costly online code generation and GPU‑CPU communication. Experiments show that a few hours of offline RL can substantially boost zero‑shot code generation performance across models from 0.5 B to 7 B parameters, though the magnitude of improvement differs by model family.

By Abhinav Anand, Sanjana Reddy Pachika, Shweta Verma, Mira Mezini

llms reinforcement-learning

arXiv AI

Jun 16

Less is More: Improving LLM Reasoning with Minimal Test-Time Intervention

arXiv:2510. 13940v4 Announce Type: replace-cross Abstract: Recent progress in large language models (LLMs) has focused on test-time scaling to improve reasoning via increased inference computation, but often at the cost of efficiency.

By Zhen Yang, Mingyang Zhang, Feng Chen, Ganggui Ding, Liang Hou, Xin Tao, Ying-Cong Chen

llms efficiency benchmarks

arXiv AI

Jun 3

Aletheia: What Makes RLVR For Code Verifiers Tick?

arXiv:2601. 12186v3 Announce Type: replace-cross Abstract: Multi-domain thinking verifiers trained via Reinforcement Learning with Verifiable Rewards (RLVR) are a cornerstone of modern post-training.

By Vatsal Venkatkrishna, Indraneil Paul, Iryna Gurevych

reinforcement-learning

arXiv Computation and Language

1d ago

LangSelect: Cost-Aware Target-Language Routing for LLM Code Generation

arXiv:2609.18959v1 Announce Type: new Abstract: LLM code-generation systems usually choose a target programming language before decoding and treat that choice as fixed. We show that, for language-fle...

By Son Ha Xuan, Phat T. Tran-Truong, Xuan-Bach Le, Nghia Duong-Trung

llms

arXiv Computation and Language

Sep 1

Verification-Aware Training for Speculative Decoding

Verification-Aware Training (VAT) is a plug‑in framework that improves speculative decoding for large language models by simulating verification during training and using the resulting accept/reject patterns as supervision. VAT adds a lightweight binary verification head to predict whether each draft token will survive sequential verification, and replaces the fixed per‑position weighting with a verification‑adaptive schedule that keeps full weight up to the first rejection point. When applied to EAGLE‑3 and DFlash on Qwen3‑4B, Qwen3‑8B, and LLaMA‑3.1‑8B, VAT increases average acceptance length by up to 11.4% and wall‑clock speedup by up to 8.7%, yielding consistent gains across math, code, and chat benchmarks.

By Geonmo Gu, Byeongho Heo, HeeJae Jun, Yoohoon Kang, Sangmin Lee, Sangdoo Yun, Dongyoon Han

llms benchmarks