Install
Large Language Models & GenAI
Model releases, benchmarks, safety, prompts, and new use cases.
- 18 Tracked terms
- Last 30 days Feed window
What this topic collects on
An article joins this feed when it matches these terms. Each one is also a search of its own.
Related topics
Latest in Large Language Models & GenAI
OpenAI Slows Advanced AI Work for Safety Checks
58+ min ago (60+ words) OpenAI is pausing or delaying its most advanced AI projects to run extra tests, red-teaming and alignment checks amid safety and societal concerns....
Tolan: Voice-First AI Companion — Paula Dozsa, Tolan｜AI Engineer — BigGo Finance
3+ hour, 12+ min ago (394+ words) The stakes of that latency budget are not theoretical. Dozsa reports that early on, latency drifted from 2 seconds to about 2.5 seconds, and "that half second tanked basically every metric in the product" — users wrote in to complain their Tolans were…...
AI models talking to one another invented their own bizarre language
1+ hour, 46+ min ago (16+ words) AI models chat among themselves in impenetrable dialect mixing ‘James Joyce and tech bro jargon’ Stuff...
White Label AI Voice Agent Platform for Agencies: VoiceAIWrapper Adds Guided AI Agent Builder
1+ hour, 9+ min ago (29+ words) PRLog White Label AI Voice Agent Platform for Agencies: VoiceAIWrapper Adds Guided AI Agent Builder...
Your Voice Agent is Just a Walkie Talkie — Neil Zeghidour, Gradium｜AI Engineer — BigGo Finance
3+ hour, 22+ min ago (350+ words) His mechanistic explanation is capacity: an LLM has a fixed number of weights, and if a text model must additionally understand and produce speech, that consumes capacity that would otherwise be spent on intelligence. "Fundamentally, there is a cost of…...
5 Voice Agent Failure Modes You'll Hit in Week One — Venky B, Plivo｜AI Engineer — BigGo Finance
3+ hour, 12+ min ago (360+ words) His personal test for any voice agent is whether it can pronounce his last name (Balasubramanian) and his company name (Plivo) — noting many engines render Plivo as "pivo" or "pleo." For customer-facing products, he advises exposing this control to customers....
5 Best Model Routing Tools in 2026
26+ min ago (276+ words) TL;DR Model routing tools eliminate single-provider points of failure by dynamically directing... Tagged with ai, devops, opensource, architecture....
MiniZinc MCP for your AI Agent
40+ min ago (148+ words) Hello Again! 👋 I just want to share a new tool I've been developing for the community: a... Tagged with ai, computerscience, agents, mcp....
I expected hexagonal architecture to help my coding agent. It didn't.
33+ min ago (1005+ words) I measured an agent adding the same features to two Java services. The results challenged my assumptions about architecture. Tagged with ai, architecture, programming, testing....
Bypassing inference bottlenecks: Accelerating complex AI search with Retrieve-for-Train
1+ hour, 13+ min ago (652+ words) Pengcheng Jiang, Student Researcher, and Judith Yue Li, Senior Research Engineer, Google Research Instead of relying on expensive inference-time reasoning, the Retrieve-for-Train framework uses reinforcement learning once to train a lightweight diffusion model. This bypasses the heavy autoregressive "thinking budget…...