INDEX

Explanations

phrases related to news headlines or bullet points

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and symbols indicating modifications in text format

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umenthal

-0.68

pill

-0.66

shift

-0.62

abal

-0.58

pex

-0.57

tsky

-0.57

 Shap

-0.56

web

-0.56

 Vish

-0.56

apesh

-0.56

POSITIVE LOGITS

Associated

0.67

NOR

0.66

 sergeant

0.65

heny

0.63

arro

0.62

 BALL

0.61

riad

0.60

 STATE

0.60

outheast

0.60

 WRITE

0.59

Activations Density 0.051%

phrases related to news headlines or bullet points

punctuation marks and symbols indicating modifications in text format

No Comments

No Known Activations