What extra content goes in an LLM model card?

medium

Answer

$In addition to standard: (1) Training data composition + filtering.$
$(2) RLHF / DPO / SFT details + reward model provenance.$
$(3) Safety evaluations (harm categories, jailbreak resistance).$
$(4) Language coverage + performance per language.$
$(5) Known refusals + limitations.$
$(6) Bias evaluations across demographics.$
$(7) Context length + tokenizer.$
$(8) Recommended use / out-of-scope.$
$(\mathrm{compute}\; + \;\mathrm{CO}_{2})$
$Examples: Llama 2 paper, GPT-4 System Card, Claude Model Card.$
$Transparency norm.$

Check yourself — multiple choice

Random
Training data + composition + filtering + SFT/RLHF/DPO details + RM provenance + safety evals (harm/jailbreak) + language coverage + refusals + bias per demo + context/tokenizer + use/out + environmental; Llama2/GPT-4/Claude examples
Just accuracy
Not real

LLM card: data + training + safety + languages + refusals + bias + env.

#llmops#safety

Practise MLOps & Data Quality

215 interview questions in this topic.

Read the lesson Flashcards Take the quiz

Related questions