INDEX

Explanations

direct speech statements made by individuals

oai_token-act-pair · gpt-3.5-turbo

quotations and reported speech

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 stub

-0.57

 harvest

-0.54

 woodland

-0.53

 fucking

-0.52

tumblr

-0.52

nig

-0.51

 pneumonia

-0.51

 broom

-0.51

 mirac

-0.50

 mole

-0.49

POSITIVE LOGITS

aturday

0.61

uer

0.59

cients

0.59

inguished

0.59

ukong

0.58

enegger

0.57

iHUD

0.56

region

0.56

bnb

0.56

emi

0.56

Activations Density 0.631%

direct speech statements made by individuals

quotations and reported speech

No Comments

No Known Activations