INDEX

Explanations

technical terms or jargon related to a specific industry, possibly related to research and product development

oai_token-act-pair · gpt-3.5-turbo

phrases indicating expert involvement or authority

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

 Eater

-0.76

llor

-0.74

agraph

-0.73

odd

-0.73

ieth

-0.71

illa

-0.70

andon

-0.69

lance

-0.68

orb

-0.68

apter

-0.68

POSITIVE LOGITS

 corridors

0.75

fav

0.70

 preference

0.70

 baths

0.64

 enrichment

0.63

bay

0.63

pak

0.62

 Palest

0.62

sts

0.61

 gardens

0.60

Activations Density 0.000%

technical terms or jargon related to a specific industry, possibly related to research and product development

phrases indicating expert involvement or authority

No Comments

No Known Activations