INDEX

Explanations

mentions of threats and potential dangers in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to various threats, particularly geopolitical or security-related threats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ricks

-0.84

urses

-0.80

arist

-0.77

ools

-0.70

uesday

-0.69

alt

-0.67

 gown

-0.67

puted

-0.66

ashion

-0.66

mys

-0.66

POSITIVE LOGITS

 posed

1.14

 threat

0.95

 threats

0.89

crow

0.86

xual

0.84

threat

0.83

 Threat

0.83

 deterrent

0.79

lessly

0.77

 glare

0.76

Activations Density 0.051%

mentions of threats and potential dangers in various contexts

references to various threats, particularly geopolitical or security-related threats

No Comments

No Known Activations