INDEX

Explanations

phrases related to misinformation, propaganda, fake news, and lies

oai_token-act-pair · gpt-3.5-turbo

references to fake news and misinformation

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

foreseen

-0.76

onen

-0.71

cit

-0.70

rontal

-0.69

ktop

-0.67

externalActionCode

-0.66

amiya

-0.65

iosyn

-0.64

airo

-0.64

emale

-0.64

POSITIVE LOGITS

 perpetrated

1.15

 debunked

1.15

 propag

1.13

 pedd

1.12

 spew

1.10

 concoct

1.06

mong

1.05

 debunk

1.05

 disinformation

1.05

 misinformation

1.04

Activations Density 0.212%

phrases related to misinformation, propaganda, fake news, and lies

references to fake news and misinformation

No Comments

No Known Activations