INDEX

Explanations

references to the location "Oregon" with varying degrees of specificity

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "gon" and its variations within a specific context

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

VIS

-0.79

\<

-0.66

SAN

-0.65

Store

-0.60

 EDITION

-0.60

EEP

-0.58

andem

-0.58

DIR

-0.58

TY

-0.57

wings

-0.57

POSITIVE LOGITS

vernment

1.22

adal

1.17

auts

1.05

osal

0.92

zo

0.92

nel

0.90

uclear

0.82

asus

0.81

sey

0.81

xon

0.80

Activations Density 0.031%

references to the location "Oregon" with varying degrees of specificity

mentions of the word "gon" and its variations within a specific context

No Comments

No Known Activations