INDEX

Explanations

references to bulls or bull-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of bulls and related terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Rad

-0.80

ä½ľ

-0.77

States

-0.76

ATIONS

-0.75

MENT

-0.74

Ô

-0.73

Privacy

-0.71

LY

-0.71

 Norn

-0.70

æĸ¹

-0.70

POSITIVE LOGITS

 bull

1.13

 bulls

1.13

 elephant

0.94

 elephants

0.91

 sharks

0.87

 horns

0.87

 shark

0.86

fights

0.81

frog

0.80

pieces

0.80

Activations Density 0.004%

references to bulls or bull-related terms

mentions of bulls and related terms

No Comments

No Known Activations