Task vectors from weight differences allow arithmetic operations to edit pre-trained models, improving multiple tasks simultaneously and enabling analogical inference on unseen tasks.
Robust fine-tuning of zero-shot models
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Randomly replacing labels in in-context demonstrations barely hurts performance, showing that label space, input distribution, and sequence format drive in-context learning more than ground-truth labels.
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
Socratic Models compose zero-shot multimodal reasoning by prompting pretrained language and vision models to exchange information and enable new capabilities without finetuning.
DG-Hard uses Donoho-Gavish hard thresholding on the fine-tuning weight delta to separate task-aligned signal from noise-like residual, recovering damaged capabilities while preserving target-task gains.
Fine-tuned small language models (3-4B parameters) preserve ordinal consistency in ranking graph structural properties for graphs larger than training data and from held-out families, showing architecture-specific degradation.
The NTIRE 2026 challenge provides a dataset of over 294,000 real and AI-generated images with 36 transformations to benchmark robust detection models.
Matched learning-rate experiments show LoRA retains substantially higher zero-shot transfer (45% vs 11% on EuroSAT, 58% vs 9% on Pets) than Full FT in CLIP adaptation.
citing papers explorer
-
Editing Models with Task Arithmetic
Task vectors from weight differences allow arithmetic operations to edit pre-trained models, improving multiple tasks simultaneously and enabling analogical inference on unseen tasks.
-
Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?
Randomly replacing labels in in-context demonstrations barely hurts performance, showing that label space, input distribution, and sequence format drive in-context learning more than ground-truth labels.
-
LAION-5B: An open large-scale dataset for training next generation image-text models
LAION-5B is an openly released dataset of 5.85 billion CLIP-filtered image-text pairs that enables replication of foundational vision-language models.
-
Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language
Socratic Models compose zero-shot multimodal reasoning by prompting pretrained language and vision models to exchange information and enable new capabilities without finetuning.
-
Spectral Unforgetting: Post-Hoc Recovery of Damaged Capabilities Without Retraining
DG-Hard uses Donoho-Gavish hard thresholding on the fine-tuning weight delta to separate task-aligned signal from noise-like residual, recovering damaged capabilities while preserving target-task gains.
-
Generalization Boundaries of Fine-Tuned Small Language Models for Graph Structural Inference
Fine-tuned small language models (3-4B parameters) preserve ordinal consistency in ranking graph structural properties for graphs larger than training data and from held-out families, showing architecture-specific degradation.
-
NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild
The NTIRE 2026 challenge provides a dataset of over 294,000 real and AI-generated images with 36 transformations to benchmark robust detection models.
-
Matched-Learning-Rate Analysis of Attention Drift and Transfer Retention in Fine-Tuned CLIP
Matched learning-rate experiments show LoRA retains substantially higher zero-shot transfer (45% vs 11% on EuroSAT, 58% vs 9% on Pets) than Full FT in CLIP adaptation.