Steering is a widely used technique for controlling large language models, yet its effects are often unstable and hard to predict. Existing theoretical accounts are largely based on the Linear Representation Hypothesis (LRH). While LRH assumes that concepts can be orthogonalized for lossless control, this idealized mapping fails in real representations and cannot account for the observed unpredict
Foundational Large Language Models (LLMs) demonstrate proficiency on a wide range of general tasks, and achieve remarkable results on various specialized tasks via domain-expert LLMs. With the ever-growing list of available LLMs, inference routers are being proposed to select the most appropriate LLM for each prompt. However, existing routing methods either optimize cost across weak-to-strong gene
Biomedical abstracts play a critical role in downstream NLP applications, such as information retrieval, biocuration, and biomedical knowledge discovery. However, a non-trivial number of biomedical articles do not have abstracts, diminishing the utility of these articles for downstream tasks. We propose DPR-BAG (Divide, Prompt, and Refine for Biomedical Abstract Generation), a training-free, zero-
Large language models (LLMs) are increasingly deployed through hosted APIs, making model extraction a practical threat to model ownership and service security. However, individual extraction queries often resemble benign requests, and existing evaluations often focus on single-query anomaly scoring or pure benign-versus-attacker user settings. We formulate model extraction monitoring as benign-cal
Policy-gradient methods usually optimize expected return, but many real world applications care about distributional properties of returns: tail risk, outlier robustness, or best-of-K discovery. We introduce OrderGrad, a family of likelihood-ratio and reparameterization gradient estimators for order-statistic objectives. OrderGrad optimizes finite-sample L-statistics, i.e., weighted averages of so
Capital is concentrating in infrastructure and frontier-model bets, with one outsized round dwarfing the rest of the day’s checks. A smaller application-layer raise shows early-stage money still reaching for product-specific wedges.
Capital is concentrating in infrastructure and frontier-model bets, with one outsized round dwarfing the rest of the day’s checks. A smaller application-layer raise shows early-stage money still reaching for product-specific wedges.
Capital is concentrating in infrastructure and frontier-model bets, with one outsized round dwarfing the rest of the day’s checks. A smaller application-layer raise shows early-stage money still reaching for product-specific wedges.
Named moves are sparse, but the day still shows institutional reshaping through promotions and founder status. The signal is less about churn than about who is being elevated into decision-making roles.
Named moves are sparse, but the day still shows institutional reshaping through promotions and founder status. The signal is less about churn than about who is being elevated into decision-making roles.
Named moves are sparse, but the day still shows institutional reshaping through promotions and founder status. The signal is less about churn than about who is being elevated into decision-making roles.
Three new scores on the same model surface a compact snapshot of capability across reasoning and knowledge tasks. The value today is in the clustering: one system, three evaluations, and a clearer read on where performance is being recorded.
Three new scores on the same model surface a compact snapshot of capability across reasoning and knowledge tasks. The value today is in the clustering: one system, three evaluations, and a clearer read on where performance is being recorded.
Three new scores on the same model surface a compact snapshot of capability across reasoning and knowledge tasks. The value today is in the clustering: one system, three evaluations, and a clearer read on where performance is being recorded.