INDEX

Explanations

references to militant groups or activities

oai_token-act-pair · gpt-3.5-turbo

words related to militant or extremist groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ebook

-0.79

ðĿ

-0.74

 Kinnikuman

-0.73

kj

-0.73

 Neighbor

-0.72

 Fargo

-0.71

eries

-0.71

RIP

-0.70

 DRAG

-0.69

uries

-0.69

POSITIVE LOGITS

 militant

1.17

 Islamist

0.87

ilitary

0.87

 milit

0.85

 insurgency

0.85

 extremist

0.84

 insurgent

0.81

 jihadist

0.80

 organisation

0.79

 uprising

0.76

Activations Density 0.005%

references to militant groups or activities

words related to militant or extremist groups

No Comments

No Known Activations