INDEX

Explanations

phrases related to legal or criminal actions

oai_token-act-pair · gpt-3.5-turbo

references to crime and legal contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NetMessage

-0.75

yssey

-0.67

staking

-0.67

 farewell

-0.65

 goodbye

-0.65

ventures

-0.63

 rever

-0.61

 anew

-0.61

PsyNetMessage

-0.61

aston

-0.60

POSITIVE LOGITS

 sexes

1.20

 slightest

1.18

 mentally

0.92

 smallest

0.90

 wearer

0.86

 highest

0.83

 same

0.83

 judiciary

0.82

 genitals

0.82

 priesthood

0.81

Activations Density 0.263%

phrases related to legal or criminal actions

references to crime and legal contexts

No Comments

No Known Activations