# Shortcut Before Circuit: Document Statistics Time In-Context Conflict Resolution

arXiv: https://arxiv.org/abs/2608.24460 · HTML: https://mlsanity.com/paper/2608.24460/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 35 of its 37 references that are in the map.
- First posted on arXiv 25 Aug 2026.

## Builds on (most cited first, 30 of 35)

- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Progress measures for grokking via mechanistic interpretability](https://mlsanity.com/paper/2301.05217/) (2023)
- [Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets](https://mlsanity.com/paper/2201.02177/) (2022)
- [Are Emergent Abilities of Large Language Models a Mirage?](https://mlsanity.com/paper/2304.15004/) (2023)
- [Underspecification Presents Challenges for Credibility in Modern Machine Learning](https://mlsanity.com/paper/2011.03395/) (2020)
- [Causal Abstractions of Neural Networks](https://mlsanity.com/paper/2106.02997/) (2021)
- [Entity-Based Knowledge Conflicts in Question Answering](https://mlsanity.com/paper/2109.05052/) (2021)
- [Adaptive Chameleon or Stubborn Sloth: Revealing the Behavior of Large Language Models in Knowledge Conflicts](https://mlsanity.com/paper/2305.13300/) (2023)
- [Data Distributional Properties Drive Emergent In-Context Learning in Transformers](https://mlsanity.com/paper/2205.05055/) (2022)
- [Physics of Language Models: Part 3.1, Knowledge Storage and Extraction](https://mlsanity.com/paper/2309.14316/) (2023)
- [Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representations](https://mlsanity.com/paper/2303.02536/) (2023)
- [Predictive Multiplicity in Classification](https://mlsanity.com/paper/1909.06677/) (2019)
- [Pretraining task diversity and the emergence of non-Bayesian in-context learning for regression](https://mlsanity.com/paper/2306.15063/) (2023)
- [Rich Knowledge Sources Bring Complex Knowledge Conflicts: Recalibrating Models to Reflect Conflicting Evidence](https://mlsanity.com/paper/2210.13701/) (2022)
- [The mechanistic basis of data dependence and abrupt learning in an in-context classification task](https://mlsanity.com/paper/2312.03002/) (2023)
- [Tug-of-War Between Knowledge: Exploring and Resolving Knowledge Conflicts in Retrieval-Augmented Language Models](https://mlsanity.com/paper/2402.14409/) (2024)
- [Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Models](https://mlsanity.com/paper/2402.18154/) (2024)
- [What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation](https://mlsanity.com/paper/2404.07129/) (2024)
- [The Transient Nature of Emergent In-Context Learning in Transformers](https://mlsanity.com/paper/2311.08360/) (2023)
- [Competition Dynamics Shape Algorithmic Phases of In-Context Learning](https://mlsanity.com/paper/2412.01003/) (2024)
- [Understanding In-Context Learning via Supportive Pretraining Data](https://mlsanity.com/paper/2306.15091/) (2023)
- [PolyPythias: Stability and Outliers across Fifty Language Model Pre-Training Runs](https://mlsanity.com/paper/2503.09543/) (2025)
- [Everything, Everywhere, All at Once: Is Mechanistic Interpretability Identifiable?](https://mlsanity.com/paper/2502.20914/) (2025)
- [Studying Large Language Model Behaviors Under Context-Memory Conflicts With Real Documents](https://mlsanity.com/paper/2404.16032/) (2024)
- [To grok or not to grok: Disentangling generalization and memorization on corrupted algorithmic datasets](https://mlsanity.com/paper/2310.13061/) (2023)
- [When can in-context learning generalize out of task distribution?](https://mlsanity.com/paper/2506.05574/) (2025)
- [Whose Facts Win? LLM Source Preferences under Knowledge Conflicts](https://mlsanity.com/paper/2601.03746/) (2026)

## Abstract (arXiv)

When a context asserts two values for one fact, a model commits to a cue -- recency, repetition, position -- but natural data rarely makes these disagree, so behavior cannot reveal which. We train 26M-parameter transformers on a synthetic language where recency and rarity are exactly coextensive, and separate them with a minimal causal edit that inverts one cue while holding the truth, token count and answer position fixed. All 75 runs reach accuracy >= 0.999, including where the trivial heuristic fails, so no held-in evaluation distinguishes them. Under intervention the per-cell readout does not replicate: 13 of 25 cells differ by more than 0.3 in sign fraction across three seeds, the largest by 0.879 against a standard error of 0.025. The construction predicts this -- coextensive rules leave the objective indifferent between them -- and the variance is ordered by how much of the optimization each comparison releases. What replicates is timing: escape from a positional shortcut with a closed-form ceiling, monotone in redundancy. Probed before that escape, attribution reverses sign in 32 of 75 runs at unchanged accuracy, and gating on circuit formation is necessary but not sufficient. The corpus fixes when a mechanism appears, not which one -- a criterion for when mechanistic attribution to data is available at all, and our construction makes the unavailable case exact.