Weekly Reads Archive

Token pruning has emerged as a mainstream approach for developing efficient Video Large Language Models (Video LLMs). This work revisits and advanc...

Large language models (LLMs) undergo alignment training to avoid harmful behaviors, yet the resulting safeguards remain brittle: jailbreaks routine...

The persistent storage requirements for high-resolution, spatiotemporally evolving fields governed by large-scale and high-dimensional partial diff...

Evidence-grounded reasoning requires more than attaching retrieved text to a prediction: a model should make decisions that depend on whether the p...

Large foundation models have made significant advances in embodied intelligence, enabling synthesis and reasoning over egocentric input for househo...

Learn how to research with ChatGPT using search and deep research to find up-to-date information, analyze sources, and generate structured insights.

Learn how finance teams use ChatGPT to streamline reporting, analyze data, improve forecasts, and communicate insights more clearly.

Learn how to use ChatGPT for research to gather sources, analyze information, and create structured, citation-backed insights.

Recent vision-language models (VLMs) typically rely on a single vision encoder trained with contrastive image-text objectives, such as CLIP-style p...

This paper explores the use of server learning for enhancing the robustness of federated learning against malicious attacks even when clients' trai...

Most of the recent generative image super-resolution (SR) methods rely on adapting large text-to-image (T2I) diffusion models pretrained on web-sca...

Non-contrast chest CTs offer a rich opportunity for both conventional pulmonary and opportunistic extra-pulmonary screening. While Multi-Task Learn...

Suppose we observe data from a distribution $P$ and we wish to test the composite null hypothesis that $P\in\mathscr P$ against a composite alterna...

OpenAI acquires TBPN to accelerate global conversations around AI and support independent media, expanding dialogue with builders, businesses, and ...

Codex now includes pay-as-you-go pricing for ChatGPT Business and Enterprise, providing teams a more flexible option to start and scale adoption.

Reconstructing 3D geometry and appearance from a sparse set of fixed cameras is a foundational task with broad applications, yet it remains fundame...

Large language model (LLM)-based coding agents achieve impressive results on controlled benchmarks yet routinely produce pull requests that real ma...

Weight tying, i.e. sharing parameters between input and output embedding matrices, is common practice in language model design, yet its impact on t...

Most recent advances in 3D generative modeling rely on diffusion or flow-matching formulations. We instead explore a fully autoregressive alternati...

Lack of accessible and dexterous robot hardware has been a significant bottleneck to achieving human-level dexterity in robots. Last year, we relea...

AI for Disaster Response in Asia: OpenAI Workshop with Gates Foundation

Learn how STADLER uses ChatGPT to transform knowledge work, saving time and accelerating productivity across 650 employees.

Video generative models show emerging reasoning behaviors. It is essential to ensure that generated events remain causally consistent across frames...

Existing tampering detection benchmarks largely rely on object masks, which severely misalign with the true edit signal: many pixels inside a mask ...

Recent advances in diffusion models have significantly improved text-to-video generation, enabling personalized content creation with fine-grained ...

Many segmentation tasks, such as medical image segmentation or future state prediction, are inherently ambiguous, meaning that multiple predictions...

Composed Video Retrieval (CoVR) aims to find a target video given a reference video and a textual modification. Prior work assumes the modification...

How OpenAI uses chain-of-thought monitoring to study misalignment in internal coding agents—analyzing real-world deployments to detect risks and st...

Recent progress in text-conditioned human motion generation has been largely driven by diffusion models trained on large-scale human motion data. B...

Machine learning approaches to spatiotemporal physical systems have primarily focused on next-frame prediction, with the goal of learning an accura...

Vision-to-code tasks require models to reconstruct structured visual inputs, such as charts, tables, and SVGs, into executable or structured repres...

Evolutions in the world, such as water pouring or ice melting, happen regardless of being observed. Video world models generate "worlds" via 2D fra...

Instruction Tuning (IT) has been proven to be an effective approach to unlock the powerful capabilities of large language models (LLMs). Recent stu...

Rakuten uses Codex, the coding agent from OpenAI, to ship software faster and safer, reducing MTTR 50%, automating CI/CD reviews, and delivering fu...

How ChatGPT defends against prompt injection and social engineering by constraining risky actions and protecting sensitive data in agent workflows.

Multimodal Large Language Models (MLLM) classification performance depends critically on evaluation protocol and ground truth quality. Studies comp...

While recent multimodal large language models (MLLMs) have made impressive strides, they predominantly employ a conventional autoregressive archite...

The integration of Large Language Models (LLMs) into autonomous driving has attracted growing interest for their strong reasoning and semantic unde...

Obstacle avoidance in unmanned aerial vehicles (UAVs), as a fundamental capability, has gained increasing attention with the growing focus on spati...

Incremental Few-Shot (IFS) segmentation aims to learn new categories over time from only a few annotations. Although widely studied in 2D, it remai...

Descript uses OpenAI models to scale multilingual video dubbing, optimizing translations for both meaning and timing so dubbed speech sounds natura...

Codex Security is an AI application security agent that analyzes project context to detect, validate, and patch complex vulnerabilities with higher...

See how Balyasny built an AI research system with GPT-5.4, rigorous model evaluation, and agent workflows to transform investment analysis at scale.

Shows that adaptively allocating test-time compute can outperform 14x larger models.

Demonstrates how RL-based training produces strong reasoning without supervised fine-tuning.

We dream of a future where point clouds from all domains can come together to shape a single model that benefits them all. Toward this goal, we pre...

Embodied Conversational Agents (ECAs) aim to emulate human face-to-face interaction through speech, gestures, and facial expressions. Current large...

Classifier-Free Guidance (CFG) has emerged as a central approach for enhancing semantic alignment in flow-based diffusion models. In this paper, we...

Many essential manipulation tasks - such as food preparation, surgery, and craftsmanship - remain intractable for autonomous robots. These tasks ar...

Achieving autonomous and versatile whole-body loco-manipulation remains a central barrier to making humanoids practically useful. Yet existing appr...

Intuitive visual walkthrough of LLM quantization techniques from FP16 to GPTQ and GGUF.

Real-world tips for efficient local LLM inference with vLLM on consumer hardware.

OpenAI introduces the Learning Outcomes Measurement Suite to assess AI's impact on student learning across diverse educational environments over time.