# Type-Safe Is Not Error-Free: Typed Decision Models Follow the Option Name, Not the Definition Bound to It

arXiv: https://arxiv.org/abs/2609.26758 · HTML: https://mlsanity.com/paper/2609.26758/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 20 references that are in the map.
- First posted on arXiv 22 Sep 2026.

## Builds on (most cited first, 11 of 11)

- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference](https://mlsanity.com/paper/2412.13663/) (2024)
- [Large Language Models Are Not Robust Multiple Choice Selectors](https://mlsanity.com/paper/2309.03882/) (2023)
- [Larger language models do in-context learning differently](https://mlsanity.com/paper/2303.03846/) (2023)
- [Mitigating Word Bias in Zero-shot Prompt-based Classifiers](https://mlsanity.com/paper/2309.04992/) (2023)
- [The Structured Output Benchmark: A Multi-Source Benchmark for Evaluating Structured Output Quality in Large Language Models](https://mlsanity.com/paper/2604.25359/) (2026)
- [Semantic Anchors in In-Context Learning: Why Small LLMs Cannot Flip Their Labels](https://mlsanity.com/paper/2511.21038/) (2025)
- [Schema-Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding](https://mlsanity.com/paper/2604.14862/) (2026)
- [Your Prompt Is Not the Only Prompt: How Much Do LLMs Weight Structured-Output Schema Descriptions?](https://mlsanity.com/paper/2608.08254/) (2026)
- [When JSON Is Not Enough: Semantic Reliability of Schema-Constrained LLM Ordering Agents](https://mlsanity.com/paper/2607.18261/) (2026)
- [PhantomFill: When the Form Demands an Answer, Language Models Invent One](https://mlsanity.com/paper/2607.20492/) (2026)

## Built on by (most cited first, 25 of 25)

- [JevAdvBench: A Benchmark and Black-Box Attacks for Reinforcement Learning for Calibrated Decisions Models](https://mlsanity.com/paper/2609.31142/) (2026)
- [Beyond Calibration: Do a Typed-Decision Model's Probabilities Obey the Probability Axioms?](https://mlsanity.com/paper/2609.33209/) (2026)
- [JevOut: Natural Context Can Flip Decision Models](https://mlsanity.com/paper/2609.30243/) (2026)
- [Typed Decision Models: An Early Evidence Audit and Evaluation Checklist](https://mlsanity.com/paper/2609.32160/) (2026)
- [Probability Contracts: Accuracy, Coherence, and Decisions Across LLM Interfaces](https://mlsanity.com/paper/2609.37470/) (2026)
- [When the Right Answer Is Missing: An Arithmetic-Dependent Rejection Bottleneck in Jev](https://mlsanity.com/paper/2609.39496/) (2026)
- [Benchmarking System One decision models against trained classifiers and language models for automated decision gates](https://mlsanity.com/paper/2610.00346/) (2026)
- [AnyJev Technical Report](https://mlsanity.com/paper/2610.00831/) (2026)
- [Labels Override Definitions in Jev-Style Typed Decision Models](https://mlsanity.com/paper/2610.02586/) (2026)
- [Benchmarking candidate coverage and rejection policy transfer in typed decision models](https://mlsanity.com/paper/2610.03387/) (2026)
- [Evaluating System One Models for Agent Security Decisions: Reliability, Calibration, and Selective Automation](https://mlsanity.com/paper/2609.33401/) (2026)
- [OmniMed-Jev: Calibrating LVLM Confidence for Trustworthy Medical Multimodal Decisions via System One](https://mlsanity.com/paper/2610.00381/) (2026)
- [SanSi: A Looped Typed Decision Model for System 1.5 Thinking](https://mlsanity.com/paper/2610.07730/) (2026)
- [Same-Number Citation Swaps: Stress-Testing Jev as a Financial Evidence Judge](https://mlsanity.com/paper/2610.08675/) (2026)
- [Adversarial Cues in Decision Models Used as Judges: The Role of Request Presentation](https://mlsanity.com/paper/2610.11436/) (2026)
- [Can Jev be Your Q or Policy in Reinforcement Learning?](https://mlsanity.com/paper/2610.11692/) (2026)
- [Specialized Decision Models vs. General-Purpose LLMs: Benchmarking Jev Across Knowledge, Reasoning, and Multilingual Tasks](https://mlsanity.com/paper/2610.11978/) (2026)
- [The Company a Question Keeps: Auditing Open Jev-Style Models for Retrieval](https://mlsanity.com/paper/2610.14631/) (2026)
- [Correct Choices, Failed Sequences: Choice2Seq Benchmarks Correctness Retention Across Output Interfaces](https://mlsanity.com/paper/2610.16191/) (2026)
- [Same Attack, Different Judge: A First Controlled Adversarial Comparison of Jev, LLM Judges, and Purpose-Built Safety Classifiers](https://mlsanity.com/paper/2610.17820/) (2026)
- [Cheap, Consistent, Calibrated, but Biased? Typed Decision Models as Judges](https://mlsanity.com/paper/2610.18122/) (2026)
- [Consider, Then Choose: Order-Invariant Decisions in System One Models](https://mlsanity.com/paper/2610.18233/) (2026)
- [How Reliable Are System-One Decisions in Medicine? A Comparative Study of Jev, Jev-like Models, and LLMs](https://mlsanity.com/paper/2610.18529/) (2026)
- [Jev-ToM: Evaluating Theory of Mind in Typed Decision Models](https://mlsanity.com/paper/2610.18610/) (2026)
- [Lost in the Fallback: How Reliably Do Typed Decision Models Use Fallback Options?](https://mlsanity.com/paper/2610.19856/) (2026)

## Abstract (arXiv)

Typed decision models return structured results, but output-type correctness alone does not ensure that decisions follow explicit option definitions. Each option pairs a name with a definition that defines its intended meaning; the name, however, can provide a competing semantic cue. We study this conflict in Jev and two open-weight models by changing only the name-definition mapping, leaving the question, state, and the names and definition texts themselves unchanged. We measure decision flips at the level of the selected definition, rather than the returned name. On 1200 decision tasks with task-specific definitions, decision-flip rates are up to 70.4 pp higher with yes/no names than with the 0/1 control. This gap holds across all 4 binary decision rules. With yes/no names, reassignment also lowers their mean AUC from 93.8% to a below-chance 23.2%. In the binary evaluations, random strings used as option names yield mean flip rates close to those of neutral controls across all three models, with comparable balanced accuracy before reassignment. Together, these results support option-name polarity as a contributor to decision instability beyond reassignment alone. The type-error rate remains 0% throughout, showing that type-correct outputs can still fail to follow explicit option definitions.