LLM stories
Large language models have moved from research curiosity to production infrastructure. This topic focuses on LLM APIs, fine-tuning, RAG pipelines, agent frameworks, and the economics of running models at scale.
HN discussions here tend to be unusually practical — benchmark comparisons, cost breakdowns, and real deployment war stories from teams shipping LLM features.
895 stories archived · Page 17 of 30
RSS feed for LLMRAGless – similar to RAG, but $0 LLM API costs at runtime
github.com · 4 points · 6 comments
The continuing "Q collar" scandal
statmodeling.stat.columbia.edu · 46 points · 15 comments
A simple fix for LLM tail latency
engineering.myhoai.com · 5 points · 0 comments
Auto-train the harness, not the LLM. cross-model, cross-benchmark gains
github.com · 4 points · 0 comments
Mirage Browser, the deadest of internets. ~1000 tok/s 1998 internet
miragebrowser.xyz · 4 points · 0 comments
How do I permanently disable random Google Photos popup to backup photos? (2024)
support.google.com · 236 points · 181 comments
Person Hides Prompt Injection in Legal Filing Telling AI to Side with Them
404media.co · 41 points · 17 comments
Virtual Private LLM, fixed fee with no usage or token limits
solheim.ai · 2 points · 0 comments
Frontier LLMs know more facts than they can recall
research.google · 10 points · 2 comments
My Grok Voice Mode System Prompt
3 points · 0 comments
ChatGPT SEO for Shopify Merchants
apps.shopify.com · 2 points · 0 comments
NanoRL – RL training for LLMs in ~1,800 lines
github.com · 10 points · 0 comments
Prompts Suno follows, not longer prompts
sunomarket.com · 3 points · 0 comments
Nine PBS could lose 70 years of archives after cloud vendor goes defunct
tomshardware.com · 98 points · 227 comments
Choosing an AI model: one prompt, 11 models, different results
netlify.com · 102 points · 95 comments
LiteLLM supply chain attack hit 2,488 organisations
twitter.com · 6 points · 0 comments
DLLM: Minimal, clean coding agent built directly on llama.cpp without overhead
github.com · 9 points · 3 comments
Decant – Understand how you spend tokens
github.com · 10 points · 0 comments
Does anyone run Postgres without PgBouncer?
brandur.org · 10 points · 113 comments
In case you think there is consciousness, intelligence or personality in an LLM [video]
youtube.com · 3 points · 8 comments
CrewScore – coverage check for AI agent prompts (0.6.11)
crewscore.ai · 3 points · 0 comments
Are there any production LLM pipeline setups to learn from?
4 points · 0 comments
Prompt Injection as Defense
arstechnica.com · 4 points · 0 comments
TokenMaxxer – Create AI spend leaderboards with your friends
tokenmaxxer.xyz · 4 points · 2 comments
Stealing Reasoning Traces from Proprietary LLM APIs
arxiv.org · 5 points · 0 comments
Research on LLM Disagreement on Factual Claims
zenodo.org · 2 points · 1 comments
Aileaks – scan repos for leaked LLM reasoning-trace secrets
github.com · 3 points · 1 comments
I took the calculator away from 50 LLMs and graded the arithmetic
github.com · 7 points · 2 comments
On Hacking – What Is Hacking?
stallman.org · 6 points · 1 comments
What sort of maths are LLMs good at?
gowers.wordpress.com · 89 points · 163 comments