Multimodal
A
Showing 1–30 of 112
-
How We Took Over ChatGPT AccountsResearchers hijack ChatGPT accounts via MathLive HTML preview flawParameter Labs found a complex attack chain against ChatGPT's sandboxed HTML previews. MathLive's virtual keyboard handles postMessage via innerHTML, letting untrusted iframe code inject HTML into the parent page and embed an unsandboxed iframe.
-
A game scripting language that runs speech-to-text 5x faster than llama.cpp on CPUGame scripting language runs speech-to-text 5x faster than llama.cpp on CPUAn article on dasllama.io describes a game scripting language that runs speech-to-text about 5x faster than llama.cpp on CPU. It was shared on Lobsters. No excerpt was available, so the language's design and the benchmark conditions are not confirmed here.
-
Latent Reasoning in Continuous Space for Unified Multimodal Models
-
RoboArchitect: Automating the Design of Robotic Agents with Persistent Memory
-
A Better Critic Matters for Distribution Matching Distillation
-
DecisionRank: On-Policy Distillation for Decision-Model Rerankers
-
Physiological Trajectories and Observation Patterns in ICU Mortality Prediction: Transportability Under Changing Measurement Conditions
-
SpiNNcloud rebrands to Sparxity, announces cluster deployment at Jülich center in GermanySpiNNcloud rebrands as Sparxity, deploys cluster at JülichSpiNNcloud has rebranded to Sparxity and announced a cluster deployment at the Jülich center in Germany. The company said the new name reflects its vision to "only compute what matters".
-
Aligning MLLMs with the Latent Structure of Human Cognition via Behavior-Derived Semantic Dimensions
-
Sensorium: An Open-Source Multimodal Data Collection and Curation Library for Embodied Surgical AI
-
Learning to Interact and Direct: Co-Evolving Video Director Agents
-
ReuScope: Auditing Data Reuse and Evaluation Sensitivity in Capsule Vision 2024
-
ElTab: Elastically Looped Tabular Transformers
-
Peeling the Matryoshka: Extracting Ontological Structure from Retrieval Embeddings
-
Multimodal In-Context Learning with Explicit Rule Induction
-
Latent-Lens: Visual Perception in Small Language Models Through Latent Communication
-
GRIT: Gradient Masking for Visually Misgrounded Tokens in Text-Rich Reasoning
-
KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents
-
From Transcription to Reformulation: Defining and Benchmarking Grounded Generative Speech Recognition
-
Rethinking Neural Thickets Without Weight Perturbations
-
Read, Allocate, Route: Role-Coupled Tokenization for Pixel-Space Generative Transformers
-
VeriCUA: Verification as Agentic Exploration for Computer-Using Agents
-
PIVOT: Progressive Intervention via Online Teacher Correction
-
Every Token Needs Its Own Teacher: Multi-teacher Adaptive On-policy Self-distillation
-
BURGER: Benchmarking and Understanding Robotic Generalization via Embodied Drawing
-
AI in the EGI e-infrastructure -- Past experiences and future strategies
-
Agentic University: Collaborative Agents for Personalized Learning
-
The Precision Wall: An Optimizer-Dependent Boundary for Optimization under Power-Law Measurement Spectra
-
Soil Moisture Estimation in Precision Agriculture: A Knowledge Distillation Approach
-
IMPACT: Inverter-Informed Multimodal Learning for Photovoltaic Generation Forecasting