From Explanations to Interventions: Execution-Guided Counterfactual Synthesis in Temporal Graphs
Not on the map. Why?
Abstract (arXiv)
Can a trace explaining model execution also compute the changes needed for a specified alternative prediction? We propose trace-guided intervention search, which uses executable reasoning traces as an intermediate representation for intervention synthesis. A Specified-Foil Counterfactual edits past events so that a frozen temporal predictor selects a designated foil. Our method constructs facts and replacement values from completed original and foil executions and recovered unmet conditions. Proposal generation constructs edits and selects candidates within a fixed cap; exact replay verifies foil top-1 outcomes among retained edits and compositions. Implemented in LiFTER for continuous-time dynamic graphs (CTDGs) and TLogic for temporal knowledge graphs (TKGs), the method improves success over coordinate-based proposal generation by 13.7-34.7 percentage points on four CTDG datasets and 60.0-83.3 points on two TKG datasets under matched downstream search and a proposal cap of 32. Separate shared-candidate comparisons retain 85.7-93.6% of black-box greedy's CTDG success rate with 75.0-80.0% fewer predictor evaluations. A Pulse case study confirms simulator-level survival for five of six interventions. Executable traces thus provide both explanatory evidence and a reusable computational representation for constructing and testing specified alternatives.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 28 references that are in the map. First posted on arXiv 10 Sep 2026.
Builds on
15 of 28 references.
References · 15
- 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,206
- 2017Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPRcited by 1,089
- 2019Predicting Dynamic Embedding Trajectory in Temporal Interaction Networkscited by 448
- 2021CF-GNNExplainer: Counterfactual Explanations for Graph Neural Networkscited by 119
- 2021TLogic: Temporal Logical Rules for Explainable Link Forecasting on Temporal Knowledge Graphscited by 59
- 2022A Survey on Graph Counterfactual Explanations: Definitions, Methods, Evaluation, and Research Challengescited by 30
- 2023TempME: Towards the Explainability of Temporal Graph Neural Networks via Motif Discoverycited by 27
- 2019Investigating Robustness and Interpretability of Link Prediction via Adversarial Modificationscited by 25
- 2023TEILP: Time Prediction over Knowledge Graphs via Logical Reasoningcited by 23
- 2023Counterfactual Learning on Graphs: A Surveycited by 22
- 2024TILP: Differentiable Learning of Temporal Logical Rules on Knowledge Graphscited by 22
- 2023Advancing Abductive Reasoning in Knowledge Graphs through Complex Logical Hypothesis Generationcited by 19
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Trace as State: Reasoning Traces as Conditional States for Long-Context Transformers
- 2025TempoBench: Reasoning Execution Without Causal Attribution Is Just Simulation
- 2025Tracing the Traces: Latent Temporal Signals for Efficient and Accurate Reasoning
- 2026Deterministic Event-Graph Substrates as World Models for Counterfactual Reasoning
- 2023GradXKG: A Universal Explain-per-use Temporal Knowledge Graph Explainer
- 2026Efficient Test-Time Scaling via Temporal Reasoning Aggregation
- 2025A New Approach to Backtracking Counterfactual Explanations: A Unified Causal Framework for Efficient Model Interpretability
- 2026LLM-Guided Knowledge Distillation for Temporal Knowledge Graph Reasoning