LLM Digest

Story

arxiv_cs_cl · May 8, 2026 · paper

Source brief

LLMs Improving LLMs: Agentic Discovery for Test-Time Scaling

arxiv.orgMay 8, 2026
original source linked

In brief

Test-time scaling (TTS) has become an effective approach for improving large language model performance by allocating additional computation during inference. However, existing TTS strategies are largely hand-crafted:...

Read the original at arxiv.org →Open in live feed

LLMs Improving LLMs: Agentic Discovery for Test-Time Scaling

Earlier in this thread 4 items

VecCISC: Improving Confidence-Informed Self-Consistency with Reasoning Trace Clustering and Candidate Answer Selection

Safety and accuracy follow different scaling laws in clinical large language models

GRAIL: A Deep-Granularity Hybrid Resonance Framework for Real-Time Agent Discovery via SLM-Enhanced Indexing

Long-Context Aware Upcycling: A New Frontier for Hybrid LLM Scaling