Phillip Isola

papers

12,875

total citations

papers (29)

DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data

NEURIPS 2023arXiv

236

citations

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

NEURIPS 2023arXiv

210

citations

Explaining in Style: Training a GAN To Explain a Classifier in StyleSpace

ICCV 2021arXiv

173

citations

Learning To Generate Line Drawings That Convey Geometry and Semantics

CVPR 2022arXiv

117

citations

Scaling Laws of Synthetic Images for Model Training ... for Now

CVPR 2024arXiv

108

citations

Any-Resolution Training for High-Resolution Image Synthesis

ECCV 2022arXiv

citations

Learning Vision from Models Rivals Learning Vision from Data

CVPR 2024arXiv

citations

Ensembling With Deep Generative Views

CVPR 2021arXiv

citations

Learning to Ground Multi-Agent Communication with Autoencoders

NEURIPS 2021arXiv

citations

Curious Representation Learning for Embodied Intelligence

ICCV 2021arXiv

citations

Persistent Nature: A Generative Model of Unbounded 3D Worlds

CVPR 2023arXiv

citations

A Vision Check-up for Language Models

CVPR 2024arXiv

citations

Procedural Image Programs for Representation Learning

NEURIPS 2022arXiv

citations

Semantic uncertainty intervals for disentangled latent spaces

NEURIPS 2022arXiv

citations

Adaptable Agent Populations via a Generative Model of Policies

NEURIPS 2021arXiv

citations

Cycle Consistency as Reward: Learning Image-Text Alignment without Human Preferences

ICCV 2025arXiv

citations

Totems: Physical Objects for Verifying Visual Integrity

ECCV 2022arXiv

citations

Personalized Representation from Personalized Generation

ICLR 2025arXiv

citations

Neural MMO 2.0: A Massively Multi-task Addition to Massively Multi-agent Learning

NEURIPS 2023

citations

Offline Multi-Agent Reinforcement Learning with Knowledge Distillation

NEURIPS 2022

citations

Learning to See by Looking at Noise

NEURIPS 2021

citations

Position: The Platonic Representation Hypothesis

ICML 2024

citations

Separating Knowledge and Perception with Procedural Data

ICML 2025

citations

Phillip Isola

papers (29)

Supervised Contrastive Learning

Contrastive Multiview Coding

What Makes for Good Views for Contrastive Learning?

Rethinking Few-shot Image Classification: A Good Embedding is All You Need?

What makes fake images detectable? Understanding properties that generalize

Improving CLIP Training with Language Rewrites

DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

Explaining in Style: Training a GAN To Explain a Classifier in StyleSpace

Learning To Generate Line Drawings That Convey Geometry and Semantics

Scaling Laws of Synthetic Images for Model Training ... for Now

Any-Resolution Training for High-Resolution Image Synthesis

Learning Vision from Models Rivals Learning Vision from Data

Ensembling With Deep Generative Views

Learning to Ground Multi-Agent Communication with Autoencoders

Curious Representation Learning for Embodied Intelligence

Persistent Nature: A Generative Model of Unbounded 3D Worlds

A Vision Check-up for Language Models

Procedural Image Programs for Representation Learning

Semantic uncertainty intervals for disentangled latent spaces

Adaptable Agent Populations via a Generative Model of Policies

Cycle Consistency as Reward: Learning Image-Text Alignment without Human Preferences

Totems: Physical Objects for Verifying Visual Integrity

Personalized Representation from Personalized Generation

Neural MMO 2.0: A Massively Multi-task Addition to Massively Multi-agent Learning

Offline Multi-Agent Reinforcement Learning with Knowledge Distillation

Learning to See by Looking at Noise

Position: The Platonic Representation Hypothesis

Separating Knowledge and Perception with Procedural Data

papers (29)

Supervised Contrastive Learning

Contrastive Multiview Coding

What Makes for Good Views for Contrastive Learning?

Rethinking Few-shot Image Classification: A Good Embedding is All You Need?

What makes fake images detectable? Understanding properties that generalize

Improving CLIP Training with Language Rewrites

DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

Explaining in Style: Training a GAN To Explain a Classifier in StyleSpace

Learning To Generate Line Drawings That Convey Geometry and Semantics

Scaling Laws of Synthetic Images for Model Training ... for Now

Any-Resolution Training for High-Resolution Image Synthesis

Learning Vision from Models Rivals Learning Vision from Data

Ensembling With Deep Generative Views

Learning to Ground Multi-Agent Communication with Autoencoders

Curious Representation Learning for Embodied Intelligence

Persistent Nature: A Generative Model of Unbounded 3D Worlds

A Vision Check-up for Language Models

Procedural Image Programs for Representation Learning

Semantic uncertainty intervals for disentangled latent spaces

Adaptable Agent Populations via a Generative Model of Policies

Cycle Consistency as Reward: Learning Image-Text Alignment without Human Preferences

Totems: Physical Objects for Verifying Visual Integrity

Personalized Representation from Personalized Generation

Neural MMO 2.0: A Massively Multi-task Addition to Massively Multi-agent Learning

Offline Multi-Agent Reinforcement Learning with Knowledge Distillation

Learning to See by Looking at Noise

Position: The Platonic Representation Hypothesis

Separating Knowledge and Perception with Procedural Data