INDEX

Explanations

names of educational institutions, especially schools

oai_token-act-pair · gpt-3.5-turbo

references to educational institutions, specifically high schools and colleges

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 strokes

-0.76

 revolutions

-0.74

 appendix

-0.70

 embodiments

-0.66

 knight

-0.65

 humour

-0.62

alys

-0.62

 invari

-0.61

 effect

-0.61

 aven

-0.61

POSITIVE LOGITS

 Complex

0.99

Ltd

0.96

 District

0.93

 Hospital

0.92

 Resort

0.91

 Facility

0.89

 Authority

0.89

 Campus

0.88

 located

0.86

itent

0.83

Activations Density 0.176%

names of educational institutions, especially schools

references to educational institutions, specifically high schools and colleges

No Comments

No Known Activations