INDEX

Explanations

the word "rude" with varying levels of intensity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "dude" and its variations

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

 ILCS

-0.95

ãĥ£

-0.82

marked

-0.80

stall

-0.77

Interstitial

-0.76

pora

-0.75

apesh

-0.74

olulu

-0.73

keep

-0.72

erald

-0.72

POSITIVE LOGITS

ude

0.77

lder

0.75

udes

0.75

xual

0.73

vich

0.69

geries

0.69

uces

0.68

cker

0.68

anu

0.68

geist

0.67

Activations Density 0.018%

the word "rude" with varying levels of intensity

instances of the word "dude" and its variations

No Comments

No Known Activations