AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Cheaply Estimating Inference Efficiency Metrics for Autoregressive Transformer Models

Deepak Narayanan, Keshav Santhanam, Peter Henderson, Rishi Bommasani, Tony Lee, Percy Liang

NeurIPS 2023 · Large language models › Language, safety and interpretability · approach: LLM unlearning and benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Send 10x more queries in parallel to the API (heavy load)Anthropic/v4-s3 raw runtime · prompt size 512end-to-end runtime: up to 2x increaseHurt
Use linear regression on a few calibration queries rather than exhaustive profilingidealized runtime estimation · 4 HELM taskscost: 57x savingsImproved
Fit a linear model of runtime in output tokens per prompt sizeopen-source models on A100 · prompt sizes 1, 512, 1024, 1536R2: > 0.999Improved
Move from A100 to V100 GPUsidealized cost · HELM tasksruntime and costHurt
Use idealized metrics instead of raw runtimesOpenAI/davinci · HELM tasks Pareto frontierPareto frontier membershipMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy