INDEX

Explanations

mentions of the word "blood"

oai_token-act-pair · gpt-3.5-turbo

references to "blood."

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

}:

-0.77

aday

-0.73

 Pinterest

-0.73

 Carrier

-0.71

reply

-0.71

 pivot

-0.71

store

-0.70

ESA

-0.70

Allows

-0.69

Works

-0.68

POSITIVE LOGITS

 blood

3.71

blood

2.56

 Blood

2.51

Blood

2.46

 saliva

1.81

 bloodstream

1.69

blooded

1.63

 urine

1.61

 semen

1.50

 bleeding

1.48

Activations Density 0.030%

mentions of the word "blood"

references to "blood."

No Comments

No Known Activations