INDEX

Explanations

phrases indicating examples or illustrations

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "say" followed by various contexts or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

hesis

-0.79

bably

-0.77

obal

-0.77

abwe

-0.74

swick

-0.73

aughs

-0.72

taboola

-0.71

xtap

-0.70

wald

-0.69

ality

-0.69

POSITIVE LOGITS

 goodbye

0.95

lihood

0.81

ings

0.74

 hello

0.72

yer

0.72

backs

0.71

ei

0.70

ership

0.64

parts

0.62

eh

0.62

Activations Density 0.034%

phrases indicating examples or illustrations

phrases that include the word "say" followed by various contexts or statements

No Comments

No Known Activations