INDEX

Explanations

phrases related to stun weapons/actions

oai_token-act-pair · gpt-3.5-turbo

references to the word "stun" and its variations in the context of violence or harm

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aternity

-0.74

yss

-0.70

ĨĴ

-0.69

 Janeiro

-0.69

å¥³

-0.68

ournal

-0.65

ļéĨĴ

-0.65

 Cors

-0.64

istle

-0.64

vernment

-0.63

POSITIVE LOGITS

ners

1.14

 stun

1.01

ned

1.00

bang

0.99

ted

0.98

ner

0.93

ting

0.89

geon

0.89

ingly

0.88

locks

0.85

Activations Density 0.009%

phrases related to stun weapons/actions

references to the word "stun" and its variations in the context of violence or harm

No Comments

No Known Activations