INDEX

Explanations

phrases related to threats and security

oai_token-act-pair · gpt-3.5-turbo

phrases indicating threats to security or safety

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

ricks

-0.77

UGC

-0.65

 Edited

-0.64

 elig

-0.64

ikes

-0.61

 reluct

-0.60

 yells

-0.60

 pitches

-0.60

ele

-0.60

 complains

-0.59

POSITIVE LOGITS

 stability

1.26

 sovereignty

1.23

 homeland

1.14

 peace

1.08

 democracy

1.05

 hegemony

1.03

 democracies

1.00

 civilization

0.99

 liberties

0.99

 liberty

0.96

Activations Density 0.218%

phrases related to threats and security

phrases indicating threats to security or safety

No Comments

No Known Activations