Social Agents: Collective Intelligence Improves LLM Predictions
In human society, collective decision making has often outperformed the judgment of individuals. Classic examples range from estimating …
PDF Cite
Social Agents: Collective Intelligence Improves LLM Predictions
ALPHA: Action-Based Learning for Pluralistic Human Alignment in Large Language Models
Large language models are widely used, but aligning them with societal values remains challenging. Current approaches often rely on …
PDF Code Dataset Project Poster
ALPHA: Action-Based Learning for Pluralistic Human Alignment in Large Language Models
BrandFusion: Aligning Image Generation with Brand Styles
While recent text-to-image models excel at generating realistic content, they struggle to capture the nuanced visual characteristics …
PDF Cite
BrandFusion: Aligning Image Generation with Brand Styles
SPRO: Improving Image Generation via Self-Play
Recent advances in diffusion models have dramatically improved image fidelity and diversity. However, aligning these models with …
PDF Cite Project
SPRO: Improving Image Generation via Self-Play
Evaluating Variance in Visual Question Answering Benchmarks
Multimodal large language models (MLLMs) have emerged as powerful tools for visual question answering (VQA), enabling reasoning and …
PDF
HIRE: Lightweight High-Resolution Image Feature Enrichment for Multimodal LLMs
The integration of high-resolution image features in modern multimodal large language models has demonstrated significant improvements …
PDF
HIRE: Lightweight High-Resolution Image Feature Enrichment for Multimodal LLMs
Learning Together to Perform Better: Teaching Small-Scale LLMs to Collaborate via Preferential Rationale Tuning
LLMs such as GPT-4 have shown a remarkable ability to solve complex questions by generating step-by-step rationales. Prior works have …
PDF Cite Code
Learning Together to Perform Better: Teaching Small-Scale LLMs to Collaborate via Preferential Rationale Tuning
EOPose: Exemplar-based object reposing using Generalized Pose Correspondences
Reposing generic objects without the use of 3D models poses a significant challenge due to the absence of a standardized pose …
PDF
EOPose: Exemplar-based object reposing using Generalized Pose Correspondences
AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment
Numerous pose-guided human editing methods have been explored by the vision community due to their extensive practical applications. …
PDF Cite Project
AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment
LLaVA Finds Free Lunch: Teaching Human Behavior Improves Content Understanding Abilities Of LLMs
Communication is defined as “Who says what to whom with what effect”. A message from a communicator generates downstream …
PDF Cite
LLaVA Finds Free Lunch: Teaching Human Behavior Improves Content Understanding Abilities Of LLMs