Two Vectors Replace In-Context Demos: Structured Task Adaptation via Embeddings
cs.CL, cs.CV, cs.LG
Submitted: 2026-10-06
Updated: 2026-10-06
Comments: Technical report
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
- Qwen2.5-VL Technical Report
- LoRA: Low-Rank Adaptation of Large Language Models
- MANTIS: Interleaved Multi-Image Instruction Tuning
- Adaptive Task Vectors for Large Language Models
- Building and better understanding vision-language models: insights and future directions
- LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models
- M$^2$IV: Towards Efficient and Fine-grained Multimodal In-Context Learning via Representation Engineering
- Decoupled Weight Decay Regularization
- Hyper-ICL: Attention Calibration with Hyperbolic Anchor Distillation for Multimodal In-Context Learning
- LLaMA: Open and Efficient Foundation Language Models
- Steering Language Models With Activation Engineering
- Understanding Reasoning in Thinking Language Models via Steering Vectors
- InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
- LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention
- Representation Engineering: A Top-Down Approach to AI Transparency
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering