INDEX

Explanations

terms related to disciplinary actions and procedures

oai_token-act-pair · gpt-3.5-turbo

terms related to disciplinary actions or consequences

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

fort

-0.72

Oo

-0.71

 Trees

-0.66

 trees

-0.66

fast

-0.65

egg

-0.65

oise

-0.65

eers

-0.64

 Eggs

-0.63

oom

-0.63

POSITIVE LOGITS

 disciplinary

3.64

ciplinary

2.01

disciplinary

1.55

 discipl

1.54

 suspensions

1.41

 discipline

1.40

 reprim

1.38

 corrective

1.38

 disciplined

1.29

 punitive

1.27

Activations Density 0.016%

terms related to disciplinary actions and procedures

terms related to disciplinary actions or consequences

No Comments

No Known Activations