AURA improves implicit-need coverage by 0.07 over ReAct baselines on a 100-query benchmark by inserting an intent inference step controlled by a gap score, while cutting probes 82% on factual tasks.
Found in the middle: Calibrating Positional Attention Bias Improves Long Context Utilization
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Inference-time attention calibration with partial strength reduces positional bias in dense retrievers on SQuAD-PosQ, FineWeb-PosQ and PosIR while preserving nDCG@10.
CoreMem replaces cosine retrieval with Fisher-Rao Riemannian matching and introduces Fisher-guided discrete token distillation for syntax-aware compression, reporting +4.51 pp open-domain and +4.17 pp temporal gains on LOCOMO and LongMemEval-S while staying inside an 8 GB VRAM budget.
citing papers explorer
-
AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents
AURA improves implicit-need coverage by 0.07 over ReAct baselines on a 100-query benchmark by inserting an intent inference step controlled by a gap score, while cutting probes 82% on factual tasks.
-
Attention Calibration for Position-Fair Dense Information Retrieval
Inference-time attention calibration with partial strength reduces positional bias in dense retrievers on SQuAD-PosQ, FineWeb-PosQ and PosIR while preserving nDCG@10.
-
CoreMem: Riemannian Retrieval and Fisher-Guided Distillation for Long-Term Memory in Dialogue Agents
CoreMem replaces cosine retrieval with Fisher-Rao Riemannian matching and introduces Fisher-guided discrete token distillation for syntax-aware compression, reporting +4.51 pp open-domain and +4.17 pp temporal gains on LOCOMO and LongMemEval-S while staying inside an 8 GB VRAM budget.