Muse
PulseAugur coverage of Muse — every cluster mentioning Muse across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
Meta's new AI detection system criticized for duplicating existing tech
Meta has developed its own AI detection system called Content Seal, which embeds invisible watermarks into AI-generated images. However, critics argue that Meta should have adopted existing solutions like Google's Synth…
-
New 'Muse' optimizers explore representation geometry for LLMs
Researchers have introduced "Muse," a novel family of optimizers designed for large language models that explores the geometric properties of parameter representations. Unlike standard Muon-style optimizers, Muse's upda…
-
Meta unveils Muse Image, emphasizing safety in generative AI
Meta has introduced Muse Image, a new generative AI model designed for image creation, as part of its broader Muse ecosystem. This model is built with a strong emphasis on safety, incorporating systematic risk evaluatio…
-
New GUARD-IT method unlearns LLMs at inference time without parameter changes
Researchers have developed a new method called GUARD-IT for inference-time machine unlearning, which aims to remove specific data's influence from large language models without altering their parameters. This technique …
-
Meta's AI Muse Faces Backlash Over Hollywood Likeness Rights
Meta's AI tool, Muse, faced significant backlash from Hollywood and its actors' union, SAG-AFTRA, over concerns regarding consent and the use of likeness rights. The tool was criticized for potentially exploiting artist…
-
Meta rolls out Muse AI image generator for apps and advertisers
Meta has launched Muse, a new AI image generation model. This model is being integrated into Meta's existing platforms, including Instagram Stories and WhatsApp. Additionally, Muse will be made available to advertisers …
-
MUSE paper repurposes diffusion model timesteps for efficient multi-task vision
Researchers have introduced MUSE, a novel parameter-free approach for multi-task dense prediction using one-step diffusion models. MUSE repurposes the fixed sinusoidal timestep embedding as an endogenous task steering s…
-
New attention mechanisms boost LLM efficiency and reduce hallucination · 10 sources tracked
Researchers are developing novel attention mechanisms to improve the efficiency and capabilities of large language models (LLMs) and multimodal large language models (MLLMs). These advancements focus on optimizing spars…
-
AI generates playable game worlds frame-by-frame
AI is advancing to a point where it can generate entire playable game worlds, frame by frame. This development is being driven by new AI models like Genie 3, Muse, and Oasis. The gaming industry is responding, with Unit…
-
New frameworks advance 3D scene generation and editing
Researchers have developed new frameworks for generating and editing 3D scenes from single images. SceneConductor decomposes the process into initialization, environment construction, and multi-agent refinement, improvi…
-
New metric reveals LLM unlearning methods fail to fully forget sensitive data
A new research paper introduces \"Leak@k\", a metric designed to evaluate the effectiveness of unlearning methods in large language models (LLMs). The study found that most current unlearning techniques fail to complete…
-
New MAAT unlearning method tackles 'Why' questions with balanced benchmark
Researchers have introduced MAAT, a novel three-phase framework for targeted machine unlearning that specifically addresses the difficulty of removing causal knowledge. Existing benchmarks are skewed, underrepresenting …
-
New MUSE benchmark evaluates Text-to-CAD models on engineering criteria
Researchers have introduced MUSE, a new benchmark designed to evaluate text-to-CAD generation models. Unlike previous benchmarks that focused on single-part models and geometric similarity, MUSE assesses complex assembl…
-
SetCon advances referring segmentation with set-level concept prediction
Researchers have introduced SetCon, a novel approach to open-ended referring segmentation that treats multiple targets as a coherent set rather than individual outputs. This method reformulates the problem as explicit s…
-
DUST releases 'Evil AI Assistants' sci-fi film compilation
DUST has released a compilation of science fiction short films titled "Evil AI Assistants," exploring various dark themes involving artificial intelligence. The collection features five distinct stories, each presenting…
-
MUSE framework resolves visual tokenization trade-offs with topological orthogonality
Researchers have introduced MUSE, a novel framework designed to resolve manifold misalignment in visual tokenization. This approach utilizes Topological Orthogonality to decouple optimization within Transformers, allowi…