INDEX

Explanations

references to items or things being created or manufactured

oai_token-act-pair · gpt-3.5-turbo

instances of the word "made."

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

lance

-0.73

rams

-0.66

gest

-0.62

bid

-0.61

rush

-0.61

asses

-0.58

orse

-0.58

ths

-0.58

urb

-0.57

rast

-0.57

POSITIVE LOGITS

 made

2.90

made

2.07

 Made

1.92

Made

1.86

 created

1.55

 produced

1.47

 brought

1.44

 formed

1.44

 kept

1.43

 makes

1.38

Activations Density 0.056%

references to items or things being created or manufactured

instances of the word "made."

No Comments

No Known Activations