Zero-Shot Confidence Estimation for Small LLMs: When Supervised Baselines Aren't Worth Training

• URL: http://arxiv.org/abs/2605.02241v3
• Date: Thu, 07 May 2026 16:11:01 GMT
• Title: Zero-Shot Confidence Estimation for Small LLMs: When Supervised Baselines Aren't Worth Training
• Abstract summary: RouteLLM-style supervised baselines are compared across three 7-8B model families and two datasets.<n>Average token log-probability, which requires no training data, matches or exceeds supervised baselines in-distribution.<n>Retrieval-conditional self-assessment, a pre-generation signal, improves over bare self-assessment by up to +0.069 AUROC at 3-10x lower latency than log-probability.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: How reliably can a small language model estimate its own correctness? The answer determines whether local-to-cloud routing-escalating queries a cheap local model cannot handle-can work without supervised training data. As inference costs dominate large language model (LLM) deployment budgets, routing most queries to a cheap local model while reserving expensive cloud calls for hard cases is an increasingly common cost-control strategy. We compare zero-shot confidence signals against RouteLLM-style supervised baselines across three 7-8B model families and two datasets (1,000 and 500 queries per model, respectively). Average token log-probability, which requires no training data, matches or exceeds supervised baselines in-distribution (Area Under the Receiver Operating Characteristic curve (AUROC) 0.650-0.714 vs. 0.644-0.676) and substantially outperforms them out-of-distribution (0.717-0.833 vs. 0.512-0.564), because it measures a property of the model's generation rather than the query distribution. This paper further proposes retrieval-conditional self-assessment, a pre-generation signal that selectively injects retrieved knowledge when similarity is high, improving over bare self-assessment by up to +0.069 AUROC at 3-10x lower latency than log-probability. A supervised baseline trained on 1,000 labeled examples never exceeds the zero-shot signal. We release all code, data, and experiment logs.

Related papers

• Distribution-First Population Simulation: Collapse, Calibration, and Recall in Non-WEIRD LLM Persona Modeling [0.0]
  Synthetic-population tools run every individual as an independent large language model (LLM) agent.<n>We show that this paradigm has a basic failure mode, and we set a distribution-first corrective against it.
  arXiv  Detail & Related papers  (2026-07-17T15:45:57Z)
• Forecasting Downstream Performance of LLMs With Proxy Metrics [37.61408486519996]
  We propose to construct proxy metrics by aggregating token-level statistics, such as entropy, top-k accuracy, and expert token rank.<n>Across three settings, our proxies consistently outperform loss- and compute-based baselines.
  arXiv  Detail & Related papers  (2026-05-18T16:17:15Z)
• ComplexMCP: Evaluation of LLM Agents in Dynamic, Interdependent, and Large-Scale Tool Sandbox [61.862814740220806]
  $textbfComplexMCP$ is a benchmark designed to evaluate agents in rigorous conditions.<n>Built on the Model Context Protocol (MCP), $textbfComplexMCP$ provides over 300 meticulously tested tools.
  arXiv  Detail & Related papers  (2026-05-11T16:20:51Z)
• Epistemic Observability in Language Models [0.0]
  We find that models report highest confidence precisely when they are fabricating.<n>We prove, under formal assumptions, that this is not a capability gap but an observational one.<n>We construct a tensor interface that escapes the impossibility by exporting computational byproducts.
  arXiv  Detail & Related papers  (2026-03-20T21:59:34Z)
• Learning More from Less: Unlocking Internal Representations for Benchmark Compression [37.69575776639016]
  We introduce REPCORE, which aligns heterogeneous hidden states into a unified latent space to construct representative coresets.<n>Experiments on five benchmarks and over 200 models show consistent gains over output-based baselines in ranking correlation and estimation accuracy.
  arXiv  Detail & Related papers  (2026-01-31T13:11:39Z)
• Learn More with Less: Uncertainty Consistency Guided Query Selection for RLVR [18.494852448006462]
  Existing RLVR algorithms require large query budgets, making annotation costly.<n>We investigate whether fewer but more informative queries can yield similar or superior performance, introducing active learning (AL) into RLVR.<n>Experiments show our method consistently outperforms random and classic AL baselines, achieving full-dataset performance while training on only 30% of the data.
  arXiv  Detail & Related papers  (2026-01-30T05:41:55Z)
• Predictive Coding and Information Bottleneck for Hallucination Detection in Large Language Models [0.8552050317027305]
  Hallucinations in Large Language Models (LLMs) remain a critical barrier to high-stakes deployment.<n>We introduce [Model Name], a hybrid detection framework that combines neuroscience-inspired signal design with supervised machine learning.
  arXiv  Detail & Related papers  (2026-01-22T05:00:21Z)
• Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs [69.2486294522259]
  BaRP is a Bandit Routing-feedback with Preferences approach that trains under the same partial-feedback restriction as deployment.<n> Framed as a contextual bandit over prompt features and a user preference vector, our method simulates an online feedback setting during training and adapts its routing decisions to each new prompt.
  arXiv  Detail & Related papers  (2025-10-08T18:24:59Z)
• SPaRFT: Self-Paced Reinforcement Fine-Tuning for Large Language Models [51.74498855100541]
  Large language models (LLMs) have shown strong reasoning capabilities when fine-tuned with reinforcement learning (RL)<n>We propose textbfSPaRFT, a self-paced learning framework that enables efficient learning based on the capability of the model being trained.
  arXiv  Detail & Related papers  (2025-08-07T03:50:48Z)
• Hard Negatives, Hard Lessons: Revisiting Training Data Quality for Robust Information Retrieval with LLMs [53.78110936809744]
  Training robust and reranker models typically rely on large-scale datasets.<n>We identify and relabel false negatives in training datasets.<n>Our training dataset and code are publicly available.
  arXiv  Detail & Related papers  (2025-05-22T17:47:57Z)
• Ranked from Within: Ranking Large Multimodal Models Without Labels [73.96543593298426]
  We show that uncertainty scores derived from softmax distributions provide a robust basis for ranking models across various tasks.<n>This facilitates the ranking of LMMs on unlabeled data, providing a practical approach for selecting models for diverse target domains without requiring manual annotation.
  arXiv  Detail & Related papers  (2024-12-09T13:05:43Z)
• How Low Can You Go? Surfacing Prototypical In-Distribution Samples for Unsupervised Anomaly Detection [48.30283806131551]
  We show that UAD with extremely few training samples can already match -- and in some cases even surpass -- the performance of training with the whole training dataset. We propose an unsupervised method to reliably identify prototypical samples to further boost UAD performance.
  arXiv  Detail & Related papers  (2023-12-06T15:30:47Z)
• Bring Your Own Data! Self-Supervised Evaluation for Large Language Models [52.15056231665816]
  We propose a framework for self-supervised evaluation of Large Language Models (LLMs) We demonstrate self-supervised evaluation strategies for measuring closed-book knowledge, toxicity, and long-range context dependence. We find strong correlations between self-supervised and human-supervised evaluations.
  arXiv  Detail & Related papers  (2023-06-23T17:59:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.