Counting Clues: A Lightweight Probabilistic Baseline Can Match an LLM
By: Furong Jia , Yuan Pu , Finn Guo and more
Potential Business Impact:
Helps doctors guess sickness better using word patterns.
Large language models (LLMs) excel on multiple-choice clinical diagnosis benchmarks, yet it is unclear how much of this performance reflects underlying probabilistic reasoning. We study this through questions from MedQA, where the task is to select the most likely diagnosis. We introduce the Frequency-Based Probabilistic Ranker (FBPR), a lightweight method that scores options with a smoothed Naive Bayes over concept-diagnosis co-occurrence statistics from a large corpus. When co-occurrence statistics were sourced from the pretraining corpora for OLMo and Llama, FBPR achieves comparable performance to the corresponding LLMs pretrained on that same corpus. Direct LLM inference and FBPR largely get different questions correct, with an overlap only slightly above random chance, indicating complementary strengths of each method. These findings highlight the continued value of explicit probabilistic baselines: they provide a meaningful performance reference point and a complementary signal for potential hybridization. While the performance of LLMs seems to be driven by a mechanism other than simple frequency aggregation, we show that an approach similar to the historically grounded, low-complexity expert systems still accounts for a substantial portion of benchmark performance.
Similar Papers
Evaluating the Use of Large Language Models as Synthetic Social Agents in Social Science Research
Artificial Intelligence
Makes AI better at guessing, not knowing for sure.
Conformal Sets in Multiple-Choice Question Answering under Black-Box Settings with Provable Coverage Guarantees
Computation and Language
Makes AI answers more trustworthy and less wrong.
Quantifying and Mitigating Selection Bias in LLMs: A Transferable LoRA Fine-Tuning and Efficient Majority Voting Approach
Computation and Language
Makes AI answer questions more fairly.