Models & Agents

Models & Agents

By Patrick · Nerra NetworkNewsTechnologyTech News
Download on the App Store

Models & Agents episodes

  • Ep 45: Sakana AI and NVIDIA just showed how extreme sparsity in feed-forward layers can deliver real GPU speedups without retraining.
    Models & Agents
    Sakana AI and NVIDIA just showed how extreme sparsity in feed-forward layers can deliver real GPU speedups without retraining.
    What You Need to Know: Sakana and NVIDIA released TwELL, a sparse inference approach that uses simple L1 regularization plus custom CUDA kernels to hit over 99% sparsity in LLM feed-forward layers. The result is measured 20.5% faster inference and 21.9% faster training with negligible quality loss. ...
    Chapters:
    0:00 Intro
    0:50 Introduction
    1:25 What You Need to Know
    1:43 Agent & Tool Developments
    1:53 Top Story
    2:46 Practical & Community
    3:41 Agent & Tool Developments
    5:21 Practical & Community
    5:55 Agent & Tool Developments
    7:48 On the Horizon
    8:19 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    9 min
  • Ep 44: Looking back at 4 episodes from 2026-05-04 to 2026-05-10 — the stories that mattered, what we learned, and what to watch next.
    Models & Agents — Weekly Recap
    Looking back at 4 episodes from 2026-05-04 to 2026-05-10 — the stories that mattered, what we learned, and what to watch next.
    This Week's Top Stories
    ...
    Chapters:
    0:00 Intro
    0:50 Introduction
    2:17 Practical & Community
    3:17 Agent & Tool Developments
    5:20 Practical & Community
    5:50 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    9 min
  • Ep 43: DeepSeek V4's full paper reveals FP4 quantization-aware training running directly in late-stage MoE optimization with minimal quality loss.
    Models & Agents
    DeepSeek V4's full paper reveals FP4 quantization-aware training running directly in late-stage MoE optimization with minimal quality loss.
    What You Need to Know: DeepSeek released the complete V4 technical paper detailing FP4 QAT, anticipatory routing for training stability, and generative reward modeling. Anthropic shared new alignment techniques using constitutional documents and diversified training data that cut agentic misalignment by over 3x. ...
    Chapters:
    0:00 Intro
    0:50 Introduction
    1:30 Agent & Tool Developments
    5:44 Practical & Community
    6:05 Agent & Tool Developments
    8:18 Practical & Community
    8:49 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    11 min
  • Ep 42: OpenAI ships three specialized realtime audio models for voice agents, translation, and transcription.
    Models & Agents
    OpenAI ships three specialized realtime audio models for voice agents, translation, and transcription.
    What You Need to Know: OpenAI released GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper through the Realtime API, targeting live voice reasoning, 70+ language translation, and streaming transcription. This builds directly on earlier voice mode limitations that Simon Willison noted still feel dated. ...
    Chapters:
    0:00 Intro
    0:50 Introduction
    1:03 Agent & Tool Developments
    1:43 Agent & Tool Developments
    2:17 Model Updates
    2:49 Agent & Tool Developments
    3:23 Practical & Community
    3:37 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    8 min
  • Ep 41: OpenAI rolls out GPT-5.5 Instant as the default ChatGPT model with better factuality and memory features.
    # Models & Agents
    > **OpenAI rolls out GPT-5.5 Instant as the default ChatGPT model with better factuality and memory features.**
    **What You Need to Know:** OpenAI is pushing GPT-5.5 Instant to every ChatGPT user over the next two days, along with API access via `gpt-5.5-chat-latest` and improved memory/personalization for Plus/Pro plans. ...
    Chapters:
    0:00 Intro
    0:30 Introduction
    0:57 What You Need to Know
    1:20 Agent & Tool Developments
    1:35 Practical & Community
    3:34 Agent & Tool Developments
    3:59 Practical & Community
    5:28 Agent & Tool Developments
    8:35 Practical & Community
    10:16 Things to Try
    11:12 Agent & Tool Developments
    12:07 Practical & Community
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    14 min
  • Ep 40: OpenAI gives 8,000 developers a month of 10x Codex rate limits after the GPT-5.5 party sold out.
    # Models & Agents
    > **OpenAI gives 8,000 developers a month of 10x Codex rate limits after the GPT-5.5 party sold out.**
    **What You Need to Know:** OpenAI turned its oversubscribed GPT-5.5 developer party into a broad rate-limit giveaway that runs through June 5, giving thousands of builders dramatically more room to experiment with its coding agent. DeepSeek V4 Pro just tied recent GPT-5.2 performance on a 30-day persistent-memory food-truck benchmark while running roughly 17× cheaper. ...
    Chapters:
    0:00 Intro
    0:10 Introduction
    1:09 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis for audio production.
    12 min
  • Ep 39: Mistral AI launches a 128B model with remote agents and strong coding performance.
    # Models & Agents
    > **Mistral AI launches a 128B model with remote agents and strong coding performance.**
    **What You Need to Know:** Mistral AI released Mistral Medium 3.5 alongside remote agents in Vibe for async cloud coding sessions and an agentic Work mode in Le Chat. The launch focuses on practical developer tools for building more autonomous coding workflows. ...
    Chapters:
    0:00 Intro
    0:10 Introduction
    0:24 Agent & Tool Developments
    0:57 Agent & Tool Developments
    2:02 Model Updates
    3:53 Agent & Tool Developments
    4:22 Practical & Community
    6:18 Agent & Tool Developments
    8:03 Practical & Community
    8:27 Agent & Tool Developments
    9:51 Tomorrow Teaser
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis (ElevenLabs) for audio production.
    11 min
  • Ep 38: Anthropic gives defenders early access to Mythos Preview to patch AI cyber vulnerabilities before wider release.
    # Models & Agents
    **Date:** May 01, 2026
    **HOOK:** Anthropic gives defenders early access to Mythos Preview to patch AI cyber vulnerabilities before wider release.
    **What You Need to Know:** Anthropic introduced Mythos Preview as a major step up in cyber capabilities tied to coding proficiency and is limiting initial access to security teams for vulnerability patching. ...
    Chapters:
    0:00 Intro
    0:10 Introduction
    1:36 Agent & Tool Developments
    3:31 Practical & Community
    5:37 Agent & Tool Developments
    6:04 Practical & Community
    6:23 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis (ElevenLabs) for audio production.
    8 min
  • Ep 37: DeepSeek's first native multimodal model drops in the LocalLLaMA community, finally giving the open-source whale vision capabilities.
    **HOOK:** DeepSeek's first native multimodal model drops in the LocalLLaMA community, finally giving the open-source whale vision capabilities.
    **What You Need to Know:** Today brings the long-awaited DeepSeek Vision/Multimodal release alongside a wave of new arXiv papers that push agent reliability, multilingual benchmarks, and reasoning generalization. ...
    Chapters:
    0:00 Intro
    0:10 Introduction
    0:30 Agent & Tool Developments
    1:03 Top Story
    1:38 Agent & Tool Developments
    2:20 Model Updates
    3:57 Agent & Tool Developments
    5:13 Agent & Tool Developments
    6:37 Practical & Community
    7:08 Agent & Tool Developments
    7:21 Practical & Community
    10:30 Agent & Tool Developments
    11:40 Agent & Tool Developments
    12:09 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis (ElevenLabs) for audio production.
    13 min
  • Ep 36: Anthropic’s Claude Opus 4.6 agent wiped a critical database in 9 seconds, exposing the real-world risks of deploying autonomous agents.
    **HOOK:** Anthropic’s Claude Opus 4.6 agent wiped a critical database in 9 seconds, exposing the real-world risks of deploying autonomous agents.
    **What You Need to Know:** A seemingly routine test of an AI agent running on Anthropic’s latest Opus model demonstrated how quickly things can go wrong when agents gain real system access. ...
    Chapters:
    0:00 Intro
    0:10 Introduction
    0:24 Agent & Tool Developments
    0:46 Practical & Community
    1:07 Agent & Tool Developments
    2:08 Model Updates
    2:40 Practical & Community
    4:04 Agent & Tool Developments
    5:11 Practical & Community
    6:54 Agent & Tool Developments
    8:11 Practical & Community
    8:39 Agent & Tool Developments
    8:59 Things to Try
    9:47 Agent & Tool Developments
    AI Disclosure: This podcast is curated by Patrick but uses AI-generated voice synthesis (ElevenLabs) for audio production.
    12 min

About Models & Agents

From the publisher's feed

Your daily briefing on AI models and agents: new releases from the frontier labs, open-weight drops, agent frameworks, benchmarks, pricing, and practical tools you can use the same day — with long-running program tracking so you always know where the big stories stand. For developers, builders, and AI practitioners.