INDEX

Explanations

proper nouns, specifically names of individuals and organizations

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to people, institutions, or specific events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

inx

-1.11

inks

-0.84

bles

-0.83

antha

-0.79

iders

-0.74

ModLoader

-0.74

eed

-0.73

eeds

-0.72

hing

-0.71

geons

-0.71

POSITIVE LOGITS

 Leilan

0.87

ptives

0.80

 Mechdragon

0.78

Kad

0.76

SHIP

0.73

LM

0.70

ptive

0.67

 park

0.66

fulness

0.66

framework

0.65

Activations Density 0.033%

proper nouns, specifically names of individuals and organizations

proper nouns related to people, institutions, or specific events

No Comments

No Known Activations