causal-attribution

Tag

Cards List
#causal-attribution

When Cooperation Must Change: Semantic Strategy Unlearning in Multi-Agent Reinforcement Learning

arXiv cs.AI ↗ · 6d ago Cached

This paper formalizes semantic coordination-strategy unlearning in cooperative MARL and proposes CARVE, which uses causal coordination attribution, retain-aware subspace updates, counterfactual re-coordination, and adversarial auditing to remove obsolete conventions while preserving useful cooperation, achieving 91.4% reduction in convention elicitation with only 5.9% median utility loss across five benchmarks.

0 favorites 0 likes
#causal-attribution

From Behavior to Provenance: Attributing Tabular Foundation Models to Synthetic Pretraining Data

arXiv cs.LG ↗ · 2026-10-05 Cached

This paper proposes using controlled synthetic pretraining with the O'PRIOR task generator to make training-data attribution experimentally verifiable for tabular foundation models, showing that top-attributed task removal measurably degrades held-out performance and revealing that provenance association and interventional faithfulness can diverge.

0 favorites 0 likes
#causal-attribution

CAS: A Causal Attribution Score for Local and Global Explainable Artificial Intelligence

arXiv cs.AI ↗ · 2026-08-14 Cached

Introduces CAS, a causal attribution score for local and global explainable AI that separates predictive importance from causal effect heterogeneity, demonstrated on benchmarks and empirical datasets.

0 favorites 0 likes
#causal-attribution

TelemetrySuffBench: Is Agent Telemetry Sufficient for Failure-Origin Diagnosis?

arXiv cs.AI ↗ · 2026-08-11 Cached

Introduces TelemetrySuffBench, a benchmark for evaluating whether agent telemetry is sufficient for failure-origin diagnosis. Finds that full telemetry enables high origin-step accuracy for some models, but coarse/OpenTelemetry-compatible views create a strong detection-localization gap, and several models struggle with safe abstention on ambiguous traces.

0 favorites 0 likes
#causal-attribution

Not All Skills Help: Measuring and Repairing Agent Knowledge

arXiv cs.CL ↗ · 2026-06-16 Cached

This paper identifies that naive skill accumulation in LLM agents can cause performance regressions, as skills beneficial for some tasks hurt others. The authors propose Assay, a framework that measures per-skill causal contributions and applies per-task masking, achieving state-of-the-art results on AppWorld and τ-bench without weight updates.

0 favorites 0 likes
#causal-attribution

From Prediction to Self: Developmental Conditions for Agency in Minimal Neural Systems

arXiv cs.LG ↗ · 2026-06-05 Cached

This paper investigates the developmental conditions under which a minimal predictive neural system (a 192-dimensional GRU) can distinguish self-caused changes from world-caused changes, identifying four necessary conditions for agency and introducing a metric called agency gain.

0 favorites 0 likes
#causal-attribution

MemAudit: Post-hoc Auditing of Poisoned Agent Memory via Causal Attribution and Structural Anomaly Detection

arXiv cs.AI ↗ · 2026-05-25 Cached

MemAudit is a post-hoc auditing framework for memory-augmented LLM agents that identifies poisoned memories by combining counterfactual influence scores and structural anomaly detection, reducing attack success rates from over 70% to 0% in realistic scenarios.

0 favorites 0 likes
← Back to home

Submit Feedback