LLMs
10 articles tagged with this topic
Why Your AI Answer Changes Every Time: The Temperature Knob You Never Touch
Ask AI the same question three times and get three different answers. The hidden parameter behind this is called Temperature. Most non-tech users neve
LLM JSON output alone is unreliable; multi-layer validation must live in the stack
JSON-only prompts don't guarantee reliable LLM output. Production must chain constraints, schema checks, retries, and monitoring together.
AI Maps Words to Vectors, Then Learns 'Position' — It Lacks Native Order Sense
8K-word tutorial explains how AI reads in two steps: map words to vectors, then teach position. Business readers skip these—yet they're key to judging
Linux 之父给 AI debug 助手打了分：能扛苦活，遇难题就喊'不可能'
Linux's latest commit message was AI-written, Torvalds confirms. His verdict: handles grunt work, cries 'impossible' on hard bugs — a structural LLM t
EasyDistill 2.0: Alibaba Cloud Puts LLM Distillation on the Assembly Line
EasyDistill 2.0 turns LLM distillation into a configurable pipeline factory — a signal that Chinese clouds are industrializing AI replication.
AI Long-Form Writing Gets Formulaic — A Developer's 'Planned Derailment' Fix
A developer fixed AI long-form writing's self-copying habit at chapter 20 with 8 triggers + 6 disruption actions. Implications stretch beyond web fict
Huolala's Memory Engineering: The Engineering Truth Behind LLM Forgetfulness
Huolala's engineering deep-dive on their self-built LLM memory system exposes a truth: however strong the model, get the memory layer wrong and your A
Million-Param GPT on Journey to the West: Demystifying LLMs Is the New Imperative
Training a million-param mini Chinese GPT on Journey to the West locally reflects the industry's urgent need to demystify the LLM black box and master
Transformer Attention Explained: The 2017 Engine Behind LLMs' Long Memory
Attention is a core LLM principle, solving AI amnesia by weighting key info. Understanding it isn't for coding—it reveals long-text limits and compute
Quoting Daniel Stenberg
cURL's lead dev reports AI -assisted security reports have improved in quality but surged in volume , consuming hours daily.