NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEED
Markets: OPENRefresh: Models tracked: Active deals: Regulatory actions: Sources:
← Back to latest

SpaceX IPO anchors a capital-heavy AI cycle

19 items · 5 desks · 10 min read
Policy Highlights5
View on Policy Monitor →

ITI and ARI Lead Coalition of over 60 Companies, Universities, and Non-Profits Calling on Congress to Enact AI Safety Institute Legislation This Year - Information Technology Industry Council (ITI)

ITI and ARI lead a coalition urging Congress to enact AI Safety Institute legislation this year. This is a call for legislative action rather than an enacted law.

POLICY HIGHLIGHTS

Inside the U.K.’s Bold Experiment in AI Safety - Time Magazine

Ambiguous signal — this appears to be a media article about the U.K.’s AI safety approach, but no specific binding regulation, consultation, or official guidance issuance is stated in the provided record.

POLICY HIGHLIGHTS

AI Safety Institute Renamed Center for AI Standards and Innovation - Broadband Breakfast

Broadband Breakfast reports that the AI Safety Institute has been renamed the Center for AI Standards and Innovation. The record does not specify any binding regulatory change or effective date.

POLICY HIGHLIGHTS

Trump admin rebrands AI safety institute in latest oversight move - CIO Dive

Ambiguous signal — a news report describes the Trump administration rebranding an AI safety institute as part of an oversight move, but the record does not specify whether any binding regulation, consultation, or formal guidance was issued.

POLICY HIGHLIGHTS

Anthropic drops Claude Gov for natsec customers; Trump administration rebrands AI Safety Institute - FedScoop

News report on Anthropic dropping Claude Gov for national security customers and a Trump administration rebranding of an AI Safety Institute. The record does not specify any binding regulation, consultation, or formal enforcement action.

POLICY HIGHLIGHTS

The Cylindrical Representation Hypothesis for Language Model Steering

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

RESEARCH

IR3DE: A Linear Router for Large Language Models

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

RESEARCH

Divide-Prompt-Refine: a Training-Free, Structure-Aware Framework for Biomedical Abstract Generation

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-

RESEARCH

An Embarrassingly Simple Detector for Model Extraction Attacks in Large Language Model API Traffic

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

RESEARCH

OrderGrad: Optimizing Beyond the Mean with Order-Statistic Policy Gradient Estimation

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

RESEARCH
SpaceX IPO anchors a capital-heavy AI cycle — Vector Wire