INDEX

Explanations

mentions of the word "aba" at varying levels of emphasis

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific name or term

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ivities

-0.80

lv

-0.76

wise

-0.75

holders

-0.72

ndra

-0.72

worthy

-0.72

rees

-0.70

rider

-0.69

etts

-0.68

cards

-0.66

POSITIVE LOGITS

aba

0.98

ipal

0.76

ãĤ§

0.76

oded

0.74

ifiable

0.73

isi

0.73

Allah

0.71

verning

0.70

uthor

0.70

ouk

0.70

Activations Density 0.022%

mentions of the word "aba" at varying levels of emphasis

mentions of a specific name or term

No Comments

No Known Activations