INDEX

Explanations

words related to political unrest and sedition

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to sedition or subversive actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

BY

-0.70

 anonymity

-0.63

 nick

-0.60

Boy

-0.59

stick

-0.57

 Soda

-0.57

 blindness

-0.55

sticks

-0.55

 Liberia

-0.54

 jurisd

-0.54

POSITIVE LOGITS

uctive

1.72

uction

1.67

entary

1.65

uct

1.51

ucing

1.50

uctions

1.48

uce

1.48

uced

1.48

uces

1.44

ucer

1.43

Activations Density 0.023%

words related to political unrest and sedition

references to sedition or subversive actions

No Comments

No Known Activations