INDEX

Explanations

strong evaluative words

oai_token-act-pair · gpt-3.5-turbo

words related to expressions of emotion or descriptors that convey disapproval

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

enhagen

-0.89

 Keller

-0.78

 Shepherd

-0.77

 Hayward

-0.75

 Jenkins

-0.73

å§«

-0.73

 Grayson

-0.73

 Frey

-0.72

 Decker

-0.71

 Lauder

-0.70

POSITIVE LOGITS

anc

1.01

withstanding

1.00

arching

0.97

rew

0.95

aring

0.95

rown

0.92

lic

0.91

ounded

0.90

arr

0.90

ob

0.89

Activations Density 0.190%

strong evaluative words

words related to expressions of emotion or descriptors that convey disapproval

No Comments

No Known Activations