INDEX

Explanations

phrases related to newspapers or classified information

oai_token-act-pair · gpt-3.5-turbo

titles and names of individuals or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 retard

-0.77

tro

-0.76

 robots

-0.73

wet

-0.71

 increment

-0.71

 doping

-0.70

 flex

-0.70

 bloc

-0.70

 undocumented

-0.70

 vigilant

-0.69

POSITIVE LOGITS

Offline

1.69

Joined

1.50

Legendary

1.50

Topic

1.49

Join

1.44

Synopsis

1.44

Enlarge

1.43

Overview

1.42

Posts

1.42

Premium

1.40

Activations Density 0.210%

phrases related to newspapers or classified information

titles and names of individuals or entities

No Comments

No Known Activations