🤖 AI Summary
This work identifies and formally names a novel phenomenon in decomposed fact-checking—termed “Decomposition-Induced Context-Memory Conflict” (DI-CC)—which arises when the decomposition stage introduces model-generated beliefs that contradict the original text, thereby producing erroneous atomic claims. The study demonstrates that while DI-CC operates via the same mechanism as classical context-memory conflict, it occurs at a distinct procedural stage. Experimental evaluations employing linear probing, SelfCheckGPT-style self-consistency sampling, and context-aware decoding reveal that linear probes effectively detect DI-CC even on unseen decompositions (AUC = 0.86–0.88), whereas self-consistency methods fail (AUC = 0.51). Although context-aware decoding mitigates DI-CC, it substantially reduces parsing accuracy on texts involving complex anaphora.
📝 Abstract
Decompose-then-verify pipelines, including FActScore-style fact-checkers and long-form factuality evaluators, first split a passage into atomic claims before checking each one. Decomposition itself is treated as a neutral preprocessing step. We show it is not: a decomposer can be induced to substitute its own parametric belief for what the source passage says, producing a claim that contradicts the text it was supposed to summarize faithfully. We call this Decomposition-Induced Context-Memory Conflict (DI-CC) and show it is mechanistically the same phenomenon as classical context-memory conflict, occurring inside a different pipeline stage than prior work has examined. A linear probe trained only on classical context-memory conflict data (NQ-Swap), never exposed to any decomposition output, significantly separates decomposition positions that produce DI-CC from faithful decompositions (AUC = 0.86-0.88, permutation p < 0.0005). An existing reference-free baseline, SelfCheckGPT-style self-consistency sampling, fails to detect DI-CC at all (AUC 0.51, chance-level), because DI-CC content is stably recoverable and recurs across resamples, unlike the variability self-consistency methods rely on. Context-aware decoding, a training-free mitigation from the classical setting, transfers to decomposition and suppresses DI-CC, but at a severe cost: many decompositions under coreference-heavy conditions fail to parse, often because the decomposer fabricates a different identity. We do not consider this mitigation deployment-ready. We further characterize the mechanism's boundaries: its natural occurrence rate is too sparss not manifest on naturally-occurring hallucinatedtext, and it requires a minimum model scale to detecablish DI-CC as a real, mechanistically grounded, andpartially treatable failure mode, with a scope we chhan overstate.