MemToC: models keep a verified-correct answer against a wrong tool in only 6.5-17.1% of cases.

Shared entity: Practical / Same source domain / Shared topic / Earlier coverage / Tension

Deep agentic search loses to a plain vector index on repo code QA. 46.2% vs 65.2%, at double the cost

Both cover Practical; reported by the same outlet (arxiv.org); overlapping topics (answer, wrong).

Shared entity: Models / Same source domain / Shared topic / Earlier coverage / Tension

Review agent refactors specifically for code that should have disappeared.

Both cover Models; reported by the same outlet (arxiv.org); overlapping topics (correct, model).

Shared entity: Models / Same source domain / Shared topic / Earlier coverage

Your CLAUDE.md holds 36.2% of the time, and that's the best case

Both cover Models; reported by the same outlet (arxiv.org); overlapping topics (against, model, tool).

Same source domain / Shared topic / Tension

Programmatic tool calling beats JSON in 11 of 14 models

Reported by the same outlet (arxiv.org); overlapping topics (against, beat, held, model, tool); pushes against this story (against).

Shared entity: Practical / Shared topic / Earlier coverage / Downstream implication

Thinking to Recall: Reasoning Unlocks Parametric Knowledge

Both cover Practical; overlapping topics (answer, correct, factual); earlier Practical coverage from 2026-03-11.

Shared entity: Models / Same source domain / Shared topic / Earlier coverage

TRAPSBench Shows Vision Models Ignore Their Own Uncertainty

Both cover Models; reported by the same outlet (arxiv.org); overlapping topics (answer, correct).

LLMs Detect Uncertainty Internally, Then Guess Anyway

Both cover Models; reported by the same outlet (arxiv.org); overlapping topics (answer, model).

Shared entity: Practical / Same source domain / Shared topic / Earlier coverage

Personalized agent skills barely beat having no skill at all

Both cover Practical; reported by the same outlet (arxiv.org); overlapping topics (against, beat).