How do you surface contradictions in an agent's knowledge before they become actions?

Contradictory memories rarely announce themselves. They surface as confident, wrong behaviour. Making disagreement visible and inspectable is a prerequisite for trusting a long-lived agent's recall.

• Contradiction detection over stored facts and their provenance
• Explanations that show which sources disagree and when they diverged
• Resolution proposals surfaced to a reviewer rather than applied silently

Evaluation in progress: precision and recall of contradiction detection on MemKit's contradiction workload.

• Detection currently assumes facts are expressible as typed assertions

Prototype running against MemKit stores; used to generate the contradiction workload labels.

• Scored detection benchmark
• Reviewer workflow for bulk resolution

• Belief revision for agent memory
• Cost of contradiction: linking stale recall to downstream action errors