PulseAugur
EN
LIVE 10:00:08

LLMs in systematic review screening: batch effects and uncertainty signals explored

By PulseAugur Editorial · [2 sources] ·

Two research papers explore the use of large language models (LLMs) in systematic review screening, a process crucial for synthesizing scientific literature. The first paper investigates class imbalance and batch effects, finding that batch processing significantly alters decision-making behavior, while prevalence metadata has limited impact. The second paper introduces auxiliary uncertainty signals from a BERT+GCN classifier to improve LLM efficiency, demonstrating that a MAYBE-only routing strategy offers the best balance of recall and cost. AI

IMPACT These studies explore methods to improve the efficiency and reliability of LLMs in scientific literature analysis, potentially accelerating research synthesis.

RANK_REASON Two academic papers published on arXiv detailing research into LLM applications for systematic reviews.

Read on arXiv cs.CL →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

LLMs in systematic review screening: batch effects and uncertainty signals explored

COVERAGE [2]

  1. arXiv cs.AI TIER_1 English(EN) · Gilberto Sussumu Hida, Danilo Monteiro Ribeiro, Clayton Suguio Hida ·

    Class Imbalance and Batch Effects in LLM-Based Screening for Systematic Reviews

    arXiv:2608.14737v1 Announce Type: cross Abstract: This study analyses LLMs in imbalanced binary classification, using study screening in systematic reviews as the application domain. An experiment was conducted in five reviews, comparing individual and batch processing, with and …

  2. arXiv cs.CL TIER_1 English(EN) · Arya Rahgozar, Pouria Mortezaagha ·

    Auxiliary uncertainty signals for LLM-assisted systematic review screening: a benchmark across eight Cohen drug-class reviews

    arXiv:2608.14551v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for title-abstract screening in systematic reviews, but their decisions lack calibrated uncertainty. We show that an auxiliary BERT+GCN classifier supplies a structured uncertainty …

RELATED ENTITIES

RELATED TOPICS