safety

arXiv:2606.06615v1 Announce Type: cross Abstract: Retrieving music using natural language descriptions has improved with contrastive audio-text models such as CLAP, but current systems remain limited

arXiv:2606.07034v1 Announce Type: new Abstract: AI-generated image detectors achieve high accuracy on in-distribution data but often fail on unseen generators. A key obstacle to understanding this fai

arXiv:2606.06885v1 Announce Type: cross Abstract: Human Image Animation has seen significant advancements, primarily driven by diffusion models. However, existing methods typically demand substantial

further evidence that people shoveling money into AI are either bad at math or blind to risk Striking paper from Wharton. The big conclusion: AI must increase productivity 2.7x -- and quickly -- or te

Gary Marcus argues that large language models based on transformers quickly reach their limitations, prompting developers to add external tools as workarounds, while neurosymbolic AI—which combines ne

arXiv:2606.07400v1 Announce Type: new Abstract: Many scientific problems require inferring unobserved mechanistic latent states from indirect observations. While classical approaches, including expect

arXiv:2606.06572v1 Announce Type: cross Abstract: We argue that modern generative models create structural risks for knowledge and cultural production at current, sub-AGI capability levels. We define

arXiv:2606.06967v1 Announce Type: new Abstract: Generative policies provide expressive and multimodal action distributions, making them attractive for reinforcement learning (RL) in complex continuous

arXiv:2606.06892v1 Announce Type: new Abstract: Scalable data attribution methods typically assign isolated utility scores to individual training examples. This prevalent additive assumption fundament

arXiv:2606.06986v1 Announce Type: new Abstract: While green finance has become a key instrument for low-carbon city transitions, its actual decarbonization effects and transmission mechanisms remain u

Historic: when the President of the United States can’t provide a single bit of specific concrete evidence to back up his serious charges, and has a temper tantrum when pressed. Trump has a meltdown a

arXiv:2606.07515v1 Announce Type: cross Abstract: We investigate the probabilistic reasoning capabilities of large language models through a controlled benchmarking study on discrete probability probl

Hypers got hype but this is an embarrassingly bad piece of analysis that sets up a false dichotomy and ignores the fact that Terence Tao also signed the Leiden Declaration. 🤦‍♂️ This post is unavailab

Gary Marcus, a prominent AI researcher and critic, is considering writing a book titled 'Seven Lies About AI' that would debunk common misconceptions about artificial intelligence, with a planned sequ

i expect value for money to increase over time as competitive pressures force providers to lower prices. but in the short time, looks like the opposite may be happening: Does a token buy you more or l

🤣 I have to think that anyone who buys the SpaceX IPO just isn’t that good at math. Or history. “Garnering that almost 600x increase means hiking sales 50% a year, on average, for a decade.” 👇🏼 https:

I will start to take Geoff Hinton seriously again (I used to and don’t anymore) when - radiologists actually start losing jobs - he acknowledges that LLMs do sometimes effectively memorize things - he

I would trust @eastdakota over Vinod Khosla any day of the week; it’s not even close. One is a man of integrity, with compassion for other human beings; the other isn’t. Uhhh… your memory is failing,

This Import AI newsletter issue covers three main topics: societal implications of reward hacking (optimizing for measurable metrics at the expense of intended goals), new reinforcement learning data

in hindsight, this was chump change New: Microsoft and OpenAI plotting new 'Stargate' supercomputer that could cost $100 billion. An incredible look at what it will take to build the next generation o

arXiv:2606.06857v1 Announce Type: new Abstract: A central goal of cognitive neuroscience is to characterize the features that are represented by human language cortex. Artificial language models (LMs)

arXiv:2606.06804v1 Announce Type: new Abstract: Digital learning environments record learners' responses to individual items, making it possible to study the development of specific skills rather than

✅ July 2024: “public sentiment towards AI will turn sharply negative. In the end, AI leaders will be viewed like cigarette industry leaders.” Prediction: @a16z’s meddling will backfire massively. US w

arXiv:2601.18510v2 Announce Type: replace-cross Abstract: While Large Language Model (LLM) agents excel at general tasks, they inherently struggle with continual adaptation due to the frozen weights a

arXiv:2606.06797v1 Announce Type: new Abstract: Cultural-aspect work on large language models is dominated by a negative target: which outputs to suppress. We argue that a constructive counterpart is

arXiv:2606.07100v1 Announce Type: new Abstract: Visual-language action (VLA) models enable robots to predict actions directly from observations and language instructions, but their performance depends

arXiv:2605.21706v2 Announce Type: replace Abstract: Safety-aligned language models are trained to refuse harmful requests, yet refusal behavior can be suppressed by steering their internal representat

arXiv:2606.06790v1 Announce Type: cross Abstract: This paper presents ERNEST, a four-wheeled planetary rover concept equipped with a two-degree-of-freedom Active Gimbal Suspension that combines yaw an

arXiv:2606.06830v1 Announce Type: cross Abstract: Data-driven pricing is increasingly prevalent in sectors such as airlines, lending, insurance, and retail. By learning demand models from customer fea

arXiv:2603.11333v2 Announce Type: replace Abstract: Short-video platforms are closed-loop, human-in-the-loop ecosystems where platform policy, creator incentives, and user behavior co-evolve. This fee

arXiv:2606.06754v1 Announce Type: cross Abstract: We present MADRAG, a training-free framework for analytic essay scoring that combines multi-agent reasoning with retrieval-augmented grounding. Unlike

many young researchers think it might be good wipe out humanity. terrifying article. who are ceding control of our planet to? Part 1 of my 'Pro-Human Manifesto' is now out! I'd love to know what you t

arXiv:2511.07380v2 Announce Type: replace Abstract: Adapting large language models (LLMs) to low-resource domains remains challenging due to the scarcity of domain-specific data. While in-domain data

arXiv:2606.06996v1 Announce Type: new Abstract: This paper studies runtime safety for autonomous driving when high-level driving commands become faulty or unreliable. Unlike conventional runtime-safet

More people need to listen to what Gary is saying here. Many, MANY of these talking heads who suck the oxygen out of the room with just their ego, say things that are so wildly wrong, that its misguid

arXiv:2606.06554v1 Announce Type: cross Abstract: Reliable polymer identification is essential for ensuring the quality and safety of recycled plastics, yet conventional sorting and spectroscopic tech

arXiv:2509.17446v3 Announce Type: replace-cross Abstract: Multimodal intent recognition (MMIR) suffers from weak semantic grounding and poor robustness under noisy or rare-class conditions. We propose

My own personal experience w Khosla is described here: The legendary investor Vinod Khosla is in the news because of questions about his integrity. I want to share my own experience: he lied viciously

arXiv:2606.07117v1 Announce Type: cross Abstract: This paper presents Native3D, the first end-to-end 3D scene generation framework that completely bypasses 2D intermediate representations. Traditional

arXiv:2606.07483v1 Announce Type: new Abstract: Many important outcomes unfold as dynamic cascades, including product adoption, disease spread, financial distress, and information diffusion. A central

arXiv:2606.06877v1 Announce Type: cross Abstract: Task planning often suffers from severe efficiency bottlenecks when robots must reason over long-horizon action sequences under complex logical constr

No. Not by itself. Sergey Brin is absolutely wrong. Transformers by themselves are not “sufficient” for AGI. Nobody uses transformers on their own anymore. Everybody is supplementing them tools and ha

Gary Marcus comments on SpaceX's valuation, suggesting that rational investors would negotiate below the asking price for the company, likely critiquing either the company's financial metrics, market

arXiv:2606.07392v1 Announce Type: new Abstract: Motivated by Large Language Model (LLM) cascading, we propose an online contextual Pandora's Box model for adaptively querying and selecting LLM APIs. I

≈ “our financials make no sense but we want to see how far we can push things” OAI files confidentially for IPO: “We have not decided on timing yet; it may be a while because there are things we want

arXiv:2606.07464v1 Announce Type: cross Abstract: Monolithic vision-action models represent an emerging paradigm in autonomous driving. However, this architecture produces token sequences that quickly