INDEX

Explanations

phrases indicating causality or logical reasoning

oai_token-act-pair · gpt-3.5-turbo

instances of the word "makes."

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

rist

-0.56

uish

-0.52

iped

-0.51

phy

-0.50

 Agency

-0.50

 Thirty

-0.50

orne

-0.49

imen

-0.49

vez

-0.49

wana

-0.49

POSITIVE LOGITS

 makes

2.83

makes

2.32

 Makes

2.16

 gives

1.93

 creates

1.90

 helps

1.82

 distinguishes

1.81

 lends

1.76

 brings

1.75

 proves

1.73

Activations Density 0.026%

phrases indicating causality or logical reasoning

instances of the word "makes."

No Comments

No Known Activations