INDEX

Explanations

mentions of the organization "Wikileaks"

oai_token-act-pair · gpt-3.5-turbo

mentions of WikiLeaks and its associated events or figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

rums

-0.77

ggle

-0.77

lasses

-0.76

omaly

-0.74

ŃĶ

-0.71

uls

-0.71

illard

-0.71

utical

-0.71

zzi

-0.70

gap

-0.70

POSITIVE LOGITS

ileaks

1.34

Leaks

1.32

 founder

1.27

 cables

1.03

 Founder

1.00

 disclosures

0.96

 WikiLeaks

0.94

 dumps

0.93

 Podesta

0.92

 revelations

0.91

Activations Density 0.046%

mentions of the organization "Wikileaks"

mentions of WikiLeaks and its associated events or figures

No Comments

No Known Activations