INDEX

Explanations

proper names such as "Norman" with varying levels of significance

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Norman"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

externalActionCode

-0.83

favorite

-0.80

Fra

-0.76

Reviewer

-0.76

Constructed

-0.74

draw

-0.72

Favorite

-0.71

plane

-0.70

EStream

-0.70

qua

-0.70

POSITIVE LOGITS

 Eisen

0.98

oland

0.84

 Bates

0.81

 Schwarz

0.79

 Orth

0.79

 Reed

0.78

 MacDonald

0.77

 Bram

0.74

brook

0.70

Orn

0.69

Activations Density 0.034%

proper names such as "Norman" with varying levels of significance

mentions of the name "Norman"

No Comments

No Known Activations