INDEX

Explanations

proper nouns, specifically the name "Bernard"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Bernard"

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

stanbul

-0.76

ngth

-0.76

worldly

-0.75

utra

-0.73

perse

-0.72

NetMessage

-0.71

pter

-0.70

licted

-0.70

conservancy

-0.69

dfx

-0.69

POSITIVE LOGITS

 Bernard

0.94

ians

0.89

ienne

0.78

ique

0.77

ello

0.75

 Suarez

0.73

ian

0.73

ardo

0.73

 Tasman

0.72

iam

0.72

Activations Density 0.014%

proper nouns, specifically the name "Bernard"

mentions of the name "Bernard"

No Comments

No Known Activations