INDEX

Explanations

names of people and places

oai_token-act-pair · gpt-3.5-turbo

names of notable individuals and their roles or professions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ahime

-0.95

angan

-0.89

 selfie

-0.80

ratom

-0.74

TPP

-0.73

adesh

-0.71

oÄŁ

-0.71

ÄŁ

-0.71

externalActionCode

-0.71

ADRA

-0.70

POSITIVE LOGITS

 Herbert

1.58

 Ernest

1.56

 Harold

1.56

 Alfred

1.51

 William

1.51

 Walter

1.51

 Frederick

1.44

 Arthur

1.43

 Theodore

1.42

 Eugene

1.42

Activations Density 0.243%

names of people and places

names of notable individuals and their roles or professions

No Comments

No Known Activations