INDEX

Explanations

mentions of the country Russia

oai_token-act-pair · gpt-3.5-turbo

mentions of Russia

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

âĢ¢âĢ¢âĢ¢âĢ¢

-0.84

eme

-0.78

ãĥ¼ãĥĨãĤ£

-0.73

Label

-0.71

Template

-0.70

Score

-0.70

sworth

-0.69

Mess

-0.69

uesday

-0.68

=/

-0.67

POSITIVE LOGITS

 annexed

1.05

 Federation

0.93

 Today

0.87

kaya

0.84

 invaded

0.79

 rall

0.78

borg

0.77

cius

0.77

rou

0.75

akov

0.74

Activations Density 0.032%

mentions of the country Russia

mentions of Russia

No Comments

No Known Activations