Anthropic tests Jupiter-v1-p ahead of its developer conference (2 minute read)
Anthropic appears to have started a fresh round of red teaming on a new internal build. The company is set to host its Code with Claude developer conference in San Francisco on May 6. The timing suggests that the model is being hardened ahead of an announcement timed to the event. The red team round is consistent with the company's responsible scaling policy, which calls for jailbreak probes and constitutional classifier stress tests before any frontier-class deployment.
|
Google is testing new Omni model for video generation (2 minute read)
Google is testing a new Omni model for video generation, potentially unifying its video and image-generation tools. The Omni model appears in Gemini's video generation UI, suggesting it might become a public product name. A launch during Google I/O 2026 is possible amidst increasing AI video competition.
|
OpenAI adds animated Pets and config imports to Codex (2 minute read)
OpenAI updated Codex with animated Pets, which appear as overlays on the screen and interact via short message bubbles. Codex also now auto-imports configuration files from other coding agents and features a new dictation dictionary to improve voice input accuracy. These updates aim to enhance Codex's usability and appeal as a comprehensive desktop application.
|
|
DeepSeek V4—almost on the frontier, a fraction of the price (4 minute read)
DeepSeek's latest preview models, DeepSeek-V4-Pro and DeepSeek-V4-Flash, are both one-million-token Mixture of Experts models. Pro has a total of 1.6 trillion parameters, with 49 billion active, while Flash has 284 billion total, with 13 billion active. DeepSeek-V4-Pro is now the largest open weights model. It is also a very cheap model to run.
|
Designing, Refining, and Maintaining Agent Skills at Perplexity (20 minute read)
Perplexity emphasizes modular Agent Skills for enhancing its frontier agent products, with specific designs and hierarchies to ensure high-quality user experiences. Unlike traditional software, Skill development prioritizes detailed, context-specific design principles where real queries and evaluations shape their necessity and content. Maintaining these Skills involves constant iteration, testing across multiple models, and prioritizing efficiency and simplicity due to the inherent 'cost' each Skill introduces.
|
|
AutoRound (GitHub Repo)
AutoRound is an advanced quantization toolkit designed for large language models and vision-language models. It achieves high accuracy at ultra-low bit widths with minimal tuning. AutoRound seamlessly works with Transformers, vLLM, SGLang, and more. It can quantize 7B models in 10 minutes on a single GPU.
|
|
Replit's Amjad Masad on the Cursor deal, fighting Apple, and why he'd rather not sell (8 minute read)
Replit's Amjad Masad highlights strong growth, nearing a billion-dollar run rate, and boasts a 300% net revenue retention rate. Unlike Cursor, which struggles with negative margins, Replit maintains gross margin positivity and appeals to non-technical users with its secure, end-to-end platform. While Masad remains committed to Replit's independence, he acknowledges open discussions with potential acquirers and expresses frustration with Apple's alleged discriminatory App Store practices, suggesting possible legal action.
|
How LLM Inference Works (8 minute read)
This post walks through the inference pipeline from tokenization and embeddings through stacked self-attention layers, then splits generation into two distinct phases on the same GPU: compute-bound prefill that processes all input tokens in parallel and memory-bound decode that emits one token at a time.
|
You Are Not Immune To Mode Collapse (8 minute read)
Mode collapse occurs when models repeatedly generate the most common outputs, leading to homogenous results, exemplified by AI generating more dogs over cats with unbalanced training data. It similarly impacts various domains like grant-making and music, as systems become increasingly specialized over time based on prior outputs and successes. To counteract, introduce variability or change external pressures to diversify and prevent over-specialization.
|
|
Become a curator for TLDR AI (3-5 hrs/week)
TLDR is looking for an engineer/researcher at a major AI lab or startup to help write for 1M+ subscribers. Our curators have been invited to Google I/O and OpenAI DevDay, scouted for Tier 1 VCs, and get early access to unreleased TLDR products. Learn more.
|
|
|
Love TLDR? Tell your friends and get rewards!
|
|
Share your referral link below with friends to get free TLDR swag!
|
|
|
|
|
Track your referrals here.
|
|
|
|
