task vector
- Papers
- 4
- Released code
- 0
- First seen
- Oct 2026
- Latest
- Nov 2026
4 papers in the last two months, against 0 in the two before.
Start here
Three papers to read first, in the order they came out- 1Distilled task updates can merge better than stronger teacher updates
On-policy distillation task vectors can complement RL teacher updates and compose more effectively across tasks.
Oct 2026 - 2AI steering vectors contain distinct functional parts good for control
Task vectors can be linearly decomposed into Task Recognition and Task Learning components for interpretable control.
Nov 2026 - 3Merging AI models improves by removing conflicting feature directions
Joint Null-Space Projection merges fine-tuned models without data, achieving 89.45% accuracy on eight CLIP tasks by removing cross-task subspace interference.
Nov 2026
How fast it is moving
Often studied with
continual learningA learning paradigm where a model sequentially acquires knowledge from multiple tasks without significantly forgetting earlier ones.1 togethermechanistic interpretabilityThe study of how neural networks compute through internal circuits, features, and causal interventions, rather than just input-output behavior.1 togetheron-policy distillationA training method where a student model learns from a teacher using its own current policy's outputs, often improving reasoning and instruction following.1 togetherThe papers
Most central to this idea first, not most recent.
- Research Labcs.LG
AI steering vectors contain distinct functional parts good for control
Task vectors can be linearly decomposed into Task Recognition and Task Learning components for interpretable control.
Keio University, RIKEN · Nov 2026
- Chinese Techcs.LG
Distilled task updates can merge better than stronger teacher updates
On-policy distillation task vectors can complement RL teacher updates and compose more effectively across tasks.
Institute of Automation, Chinese Academy of Sciences, University of Chinese Academy of Sciences · Oct 2026
- Academiccs.CV
Merging AI models improves by removing conflicting feature directions
Joint Null-Space Projection merges fine-tuned models without data, achieving 89.45% accuracy on eight CLIP tasks by removing cross-task subspace interference.
Tohoku University · Nov 2026
- Academiccs.LG
Language models can transfer specialized skills across sizes without retraining
Fine-tuning effects transfer across Transformers of different depths and widths by aligning activation changes and synthesizing target updates with ridge regression.
University of Modena and Reggio Emilia, University of Florence · Oct 2026
Concepts are extracted from each paper and reused across the corpus, so this page grows on its own as the desk reads.