INDEX

Explanations

phrases related to individual professions or accomplishments

oai_token-act-pair · gpt-3.5-turbo

references to career milestones or experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Lumpur

-0.71

 ease

-0.65

 liquids

-0.58

thia

-0.57

egg

-0.57

Ing

-0.56

 liquidity

-0.56

ractive

-0.56

perse

-0.56

 Grimes

-0.55

POSITIVE LOGITS

 abroad

0.83

 overseas

0.81

span

0.80

 predec

0.79

wcs

0.79

enegger

0.79

ixtape

0.77

advertising

0.77

é¾įåĸļå£«

0.74

CTV

0.73

Activations Density 0.019%

phrases related to individual professions or accomplishments

references to career milestones or experiences

No Comments

No Known Activations