INDEX

Explanations

words related to terrorism

oai_token-act-pair · gpt-3.5-turbo

references to terrorism and related subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

gling

-0.77

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.75

dit

-0.75

abase

-0.70

ixed

-0.69

flex

-0.64

ðĿ

-0.64

atur

-0.64

galitarian

-0.63

 Jaguar

-0.63

POSITIVE LOGITS

ophobic

1.01

ophobia

0.99

 Terrorism

0.97

 terrorism

0.96

 prevention

0.91

ISM

0.91

 perpetrated

0.90

 threats

0.89

 suspects

0.89

 offences

0.88

Activations Density 0.042%

words related to terrorism

references to terrorism and related subjects

No Comments

No Known Activations