Transformers Remember First, Forget Last: Dual-Process Interference in LLMs
arXiv 2603.00270•81daf13397f0f04e960a53b764107d60016d8c6fec59c484b153d23093eb994e
CoCoALLM memoryLoRAMuonRecPhotoBench benchmarkQuaSIDRAIEReFeed datasetSODASemantic IDsTARSETiny-Critic RAGagentic RAGclinical QAgenerative recommendationmultimodal embeddingsoptimizer (Muon)parameter-efficient modelsproactive interferencequery rewritingrecommender systemsretrieval-augmented generationretroactive interferencetest-time adaptationtransformers
Paper metadata
- arXiv ID
- 2603.00270
- Version
- Not specified by this published record
- Category
- Computer Science — Information Retrieval (cs.IR)
The PDF link points to arxiv.org. Baitaphish does not expose a private stored PDF.
Evidence and limitations
- Source ID
- arxiv_cs_ir
- Record identifier
- 81daf13397f0f04e960a53b764107d60016d8c6fec59c484b153d23093eb994e
- Enrichment time
- 2026-03-04T19:50:14Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.