Cast
PulseAugur coverage of Cast — every cluster mentioning Cast across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New CAST method improves graph Cholesky factorization
Researchers have introduced CAST (Canonical Approximate Schur Tree), a novel method for constructing approximate Cholesky factorizations on graphs. This technique aims to improve the efficiency of solving systems of lin…
-
New methods enhance LLM steering for behavior control · 2 sources tracked
Two new research papers introduce novel methods for steering large language models to suppress undesired behaviors. GAPS (Gated Activation steering via Posterior and Separability) employs dimension-level gates to select…
-
New CAST framework enhances audibility of clinical AI models
Researchers have developed a new framework called CAST (Concept-guided Artifact Suppression Tuning) to make clinical language models more auditable and robust. This method uses Sparse Autoencoders to identify and suppre…
-
WAN vs. LAN: Understanding Network Port Differences
The distinction between WAN (Wide Area Network) and LAN (Local Area Network) ports on networking equipment is crucial for proper internet connectivity. The WAN port connects a local network to the internet, typically vi…
-
New CAST framework enables zero-shot classifier extension without target data
Researchers have introduced CAST, a novel framework designed to extend pre-trained classifiers to new categories without requiring any examples from the target distribution. This training-free and image-free approach ut…
-
New CausalSteward framework aids causal discovery with multi-agent approach
Researchers have introduced CausalSteward (CAST), a new human-in-the-loop framework designed to help assemble large causal models from high-dimensional data. This multi-agent system employs a divide-and-conquer strategy…
-
New defense probes LLM hidden states to block prefilling attacks
Researchers have developed a new defense mechanism for large language models called response-time probing, which effectively counters prefilling attacks. This method, when combined with existing techniques like AlphaSte…
-
New research tackles LLM and VLM hallucinations with novel detection methods · 7 sources tracked
Researchers are developing new methods to combat hallucinations in large language models (LLMs) and vision-language models (VLMs). One approach, InnerExpert, leverages internal signals from Mixture-of-Experts (MoE) arch…
-
New response-time probing method boosts LLM safety against prefilling attacks
Researchers have developed a new method called response-time probing to enhance the safety of large language models by detecting prefilling attacks. This technique, which probes the model's hidden state at the first gen…
-
New CAST method improves LLM reasoning via self-distillation
Researchers have developed CAST, a novel self-distillation method designed to enhance reinforcement learning with verifiable rewards (RLVR) in large language models, particularly for Group Relative Policy Optimization (…
-
New algorithm optimizes HIV prevention resource allocation
Researchers have developed a new algorithm called CAST to optimize the distribution of resources for HIV prevention. This algorithm aims to minimize new infections by strategically treating individuals who are virally u…
-
New CAST method forecasts distribution-valued time series
Researchers have introduced CAST, a novel method for forecasting distribution-valued time series, which are observed as aggregate distributions rather than simple scalar trajectories. This approach is designed to operat…
-
AI enhances recommendation systems by extracting sensory data and modeling semantic transitions
Researchers have developed new methods for sequential recommendation systems that leverage rich semantic information from product reviews and item attributes. One approach, ASER, uses a fine-tuned large language model t…