positional-bias

Tag

Cards List
#positional-bias

Mitigating Position Bias in Transformers via Layer-Specific Positional Embedding Scaling

arXiv cs.CL · 2026-06-29 Cached

Introduces LPES, a layer-specific positional embedding scaling method that mitigates the 'lost-in-the-middle' problem in LLMs by assigning distinct scaling factors per layer using a genetic algorithm with Bézier curves, achieving up to 11.2% accuracy gain without fine-tuning or latency increase.

0 favorites 0 likes
#positional-bias

LLM Judges Have Dark Current: A Psychometric Datasheet for LLM-as-a-Judge Evaluation

arXiv cs.CL · 2026-06-16 Cached

This paper introduces a psychometric datasheet protocol for evaluating LLM judges as measurement instruments, measuring dark current, positional false preference, stable cross-sensitivity, and target sensitivity. A case study on three open-weight models reveals significant differences in judge quality and behavior.

0 favorites 0 likes
#positional-bias

A Systematic Evaluation of Positional Bias in Multi-Video Summarization with MLLMs

arXiv cs.CL · 2026-06-04 Cached

Researchers from Jilin University systematically evaluate positional bias in multi-video summarization using MLLMs, constructing a benchmark from ActivityNet and News videos and assessing nine models with metrics including Coverage, Directional Positional Bias, and Middle-Edge Gap. Results show positional effects are domain- and model-dependent, and increasing visual or generation budget does not uniformly resolve the imbalance.

0 favorites 0 likes
#positional-bias

The Readout Shortcut: Positional Number Copying Dominates Arithmetic CoT Readout in Small Language Models

arXiv cs.LG · 2026-05-25 Cached

This paper identifies a 'positional copying' shortcut where small language models answer arithmetic questions by copying the last number before the answer delimiter, bypassing actual reasoning. This effect explains why shuffling CoT steps retains performance; it accounts for 89-92% of teacher-forcing accuracy in 1-3B models on GSM8K.

0 favorites 0 likes
#positional-bias

Self-Describing Structured Data with Dual-Layer Guidance: A Lightweight Alternative to RAG for Precision Retrieval in Large-Scale LLM Knowledge Navigation

arXiv cs.CL · 2026-04-23 Cached

SDSR proposes lightweight self-describing structured data with dual-layer guidance to exploit LLM primacy bias, achieving 100% routing accuracy without vector DBs.

0 favorites 0 likes
← Back to home

Submit Feedback