INDEX

Explanations

words related to legal terms and court findings

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Ad" or related advertising terminology

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

 chops

-0.78

 fractured

-0.72

 mushroom

-0.72

 mushrooms

-0.71

 shale

-0.68

 leve

-0.67

 looting

-0.66

 mash

-0.66

 fracture

-0.64

 pole

-0.63

POSITIVE LOGITS

Ad

3.56

Ad

2.20

ad

1.54

AD

1.48

ad

1.45

Ab

1.36

Adv

1.32

Ads

1.31

Spons

1.28

Aff

1.24

Activations Density 0.009%

words related to legal terms and court findings

occurrences of the word "Ad" or related advertising terminology

No Comments

No Known Activations