DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
DeepSeek's Flash outpaced its own flagship. The upgrade was post-training, not parameters.

DeepSeek's Flash outpaced its own flagship. The upgrade was post-training, not parameters.

1
Comments
2 min read
GPT-4o-Level Vision on a Laptop? The Open-Weight Multimodal LLMs That Changed the Game

GPT-4o-Level Vision on a Laptop? The Open-Weight Multimodal LLMs That Changed the Game

Comments
5 min read
Building a Ride-Share Zone-Balancing Agent with LangGraph — Part 2: Teaching the Agent to Read Ops Notes

Building a Ride-Share Zone-Balancing Agent with LangGraph — Part 2: Teaching the Agent to Read Ops Notes

Comments
7 min read
GPT-4o mini, OpenAI's Cost-Optimized Small Model Slashing API Prices for High-Volume Workloads

GPT-4o mini, OpenAI's Cost-Optimized Small Model Slashing API Prices for High-Volume Workloads

Comments
4 min read
Node.js Support Ticket Classification with LLM JSON Schema Tags

Node.js Support Ticket Classification with LLM JSON Schema Tags

Comments
7 min read
Play in GenAI is now in its applications

Play in GenAI is now in its applications

Comments
1 min read
LLM, que bisho Ă© esse?

LLM, que bisho Ă© esse?

2
Comments 1
4 min read
Selecting a Compatible API Gateway: Token Cost, Batch Work, and Data Residency

Selecting a Compatible API Gateway: Token Cost, Batch Work, and Data Residency

Comments
7 min read
Why We Calculate Astrology Charts Before Asking an LLM to Explain Them

Why We Calculate Astrology Charts Before Asking an LLM to Explain Them

Comments
2 min read
We judged the same frozen answers 15 times. The judge wasn't the noisy part.

We judged the same frozen answers 15 times. The judge wasn't the noisy part.

Comments
6 min read
Looped reasoning means the AI's visible trace isn't the reasoning

Looped reasoning means the AI's visible trace isn't the reasoning

Comments 3
2 min read
Forensic Receipts: From Trusted to Proven

Shifts from mere trust to cryptographic proof

Forensic Receipts: From Trusted to Proven

16
Picked as gem Comments 10
3 min read
Claude Fable 5.1 + GPT-6 Astra: Two Coding Agents, One Workflow

Claude Fable 5.1 + GPT-6 Astra: Two Coding Agents, One Workflow

7
Comments 3
9 min read
Interactive AI Eval Dashboards with Data Studio

Interactive AI Eval Dashboards with Data Studio

10
Comments 2
11 min read
We beat mem0 on LongMemEval-S retrieval (+11.6pt P@1, full 500) with a fully local memory layer — no LLM at write time

We beat mem0 on LongMemEval-S retrieval (+11.6pt P@1, full 500) with a fully local memory layer — no LLM at write time

1
Comments 4
4 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.