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

arXiv: https://arxiv.org/abs/2608.14551 · HTML: https://mlsanity.com/paper/2608.14551/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 17 of its 28 references that are in the map.
- First posted on arXiv 3 May 2026.

## Builds on (most cited first, 17 of 17)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles](https://mlsanity.com/paper/1612.01474/) (2016)
- [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022)
- [Modeling Relational Data with Graph Convolutional Networks](https://mlsanity.com/paper/1703.06103/) (2017)
- [Domain-Specific Language Model Pretraining for Biomedical Natural Language Processing](https://mlsanity.com/paper/2007.15779/) (2020)
- [Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs](https://mlsanity.com/paper/2306.13063/) (2023)
- [Graph Convolutional Networks for Text Classification](https://mlsanity.com/paper/1809.05679/) (2018)
- [Tool Learning with Foundation Models](https://mlsanity.com/paper/2304.08354/) (2023)
- [OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts](https://mlsanity.com/paper/2205.01833/) (2022)
- [BertGCN: Transductive Text Classification by Combining GCN and BERT](https://mlsanity.com/paper/2105.05727/) (2021)
- [Can large language models replace humans in the systematic review process? Evaluating GPT-4's efficacy in screening and extracting data from peer-reviewed and grey literature in multiple languages](https://mlsanity.com/paper/2310.17526/) (2023)
- [Open Source Software for Efficient and Transparent Reviews](https://mlsanity.com/paper/2006.12166/) (2020)
- [Automated Paper Screening for Clinical Reviews Using Large Language Models](https://mlsanity.com/paper/2305.00844/) (2023)
- [Zero-shot Generative Large Language Models for Systematic Review Screening Automation](https://mlsanity.com/paper/2401.06320/) (2024)

## Abstract (arXiv)

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 signal that improves LLM screening efficiency, and we identify the prompt-delivery strategy that maximises the benefit-to-cost ratio. We evaluate five LLM prompt-delivery conditions on eight drug-class datasets from the Cohen (2006) benchmark using 3 seeds x 5-fold stratified cross-validation (600 fold-level results). A BERT+GCN model trained per fold classifies each test paper as INCLUDE, EXCLUDE, or MAYBE via two spectral tests (algebraic radical and categorical paradox). Conditions vary information content (none / label / full scores), selectivity (all papers vs. MAYBE only), and timing (proactive vs. reactive two-pass). A cross-model pilot against gpt-4.1-mini on three datasets tests cross-generation transfer. Three findings: (i) Full-context delivery yields significant gains in F1 (+0.011, paired Wilcoxon p=0.008) and WSS@95 (+0.050, p=0.039) at a 1.28x token-cost premium, while preserving recall. (ii) MAYBE-only routing is Pareto-optimal: highest mean recall (0.92) and AUC-ROC (0.54) at only 1.05x baseline cost -- one sixth of full-context overhead. (iii) The two-pass design escalates 22.2% +/- 8.8% of records yet never revises its decision (0% flip rate across all datasets and folds), giving decisive evidence that current instruction-tuned LLMs cannot self-triage. The cross-model pilot shows an identical +0.8% recall uplift for both LLM generations. A per-paper ablation across 20,796 observations shows the dual paradox test reduces empirically to a one-line logit-gap criterion. We release the full pipeline; the 600-run experiment replays in under one hour from cached LLM responses.