Menu

#Tokens

477 posts

Feed·
12 of 477 posts
How do you generate embeddings for millions of documents?
🖼️
0Read More

How do you generate embeddings for millions of documents?

DEV Community: datascience·Viswaretas Kotra·about 2 hours ago
#kmoDm8wk
#devtodatascience#science#devto#dev#model

Shard the corpus, run GPU workers that skip finished shards after a restart, and do the math: 10M documents cost $6 to $11 of GPU time vs $102 by API.

15s
SCIM Provisioning & Deprovisioning That Actually Removes Access
📰
0Read More

SCIM Provisioning & Deprovisioning That Actually Removes Access

DEV Community: authentication·Muhammad Ali Mughal·about 2 hours ago
#2X6cIzJ7
#devtoauthentication#devto#dev#user#scim

SSO stops new logins, but it does not end the sessions, API tokens and invites a departed user already has. How I build a SCIM 2.0 endpoint in a B2B SaaS app so that deactivating a user in Okta or Entra ID really removes access: the PATCH active=false…

15s
Escaping Sunk-Cost Retry Loops in Coding Agents: Single-Session Retries vs. a Parallel Best-of-N Tournament (-67.2% Input Tokens)
📰
0Read More

Escaping Sunk-Cost Retry Loops in Coding Agents: Single-Session Retries vs. a Parallel Best-of-N Tournament (-67.2% Input Tokens)

DEV Community: ai·tappers·about 8 hours ago
#huuNdpEO
#devtoai#devto#dev#branch#tokens

💡 AI Labo Benchmark Series — Experiment E13 (Parallel Worktrees & Best-of-N Selection) This...

15s
Why LLMs Fail at Unified Diffs: Standard Diff Hunks vs. Line-Number-Free SEARCH/REPLACE Blocks (2.41x Faster)
📰
0Read More

Why LLMs Fail at Unified Diffs: Standard Diff Hunks vs. Line-Number-Free SEARCH/REPLACE Blocks (2.41x Faster)

DEV Community: ai·tappers·about 8 hours ago
#t5Y37sxN
#devtoai#devto#dev#line#tokens

💡 AI Labo Benchmark Series — Experiment E15 (Edit Primitives & Tool Ergonomics) This post is...

15s
Token price lies: how Gemini 4 Neon cuts your API bill
🖼️
0Read More

Token price lies: how Gemini 4 Neon cuts your API bill

DEV Community: api·Andrey Altrouter·about 13 hours ago
#wXjpzRvo
#devtoapi#devto#dev#tokens#model

Switching coding agents to Gemini 4 Neon saves up to 36% of your budget through free cache writes and lower reasoning levels. The math of hidden tokens.

15s
Design Tokens for AI: Extract a Website's Real Colours, Fonts, Shadows and Icons So an LLM Builds On Brand
📰
0Read More

Design Tokens for AI: Extract a Website's Real Colours, Fonts, Shadows and Icons So an LLM Builds On Brand

DEV Community: css·Sandeep Virk·about 13 hours ago
#LhJwKhHi
#devtocss#devto#dev#tokens#model

Design tokens for AI: pull a site's real colours, gradients, fonts, shadows and icons, turn them into tokens and get on-brand output from Claude.

15s