INDEX

Explanations

words related to legal offenses and criminal behavior

oai_token-act-pair · gpt-3.5-turbo

terms related to sports and legal wrongdoing specifically through the use of the word "offense."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

AppData

-0.77

 modelling

-0.74

Thom

-0.67

Virgin

-0.66

wen

-0.66

mini

-0.65

ween

-0.64

tree

-0.64

 Clover

-0.63

 Thomson

-0.63

POSITIVE LOGITS

 offense

4.06

 offence

2.99

 Offense

2.74

 offenses

2.57

 defense

1.83

 offensive

1.70

 offensively

1.65

 Offensive

1.46

defense

1.43

offensive

1.38

Activations Density 0.009%

words related to legal offenses and criminal behavior

terms related to sports and legal wrongdoing specifically through the use of the word "offense."

No Comments

No Known Activations