INDEX

Explanations

negative opinions or criticism

oai_token-act-pair · gpt-3.5-turbo

dash-separated phrases related to various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Dickinson

-0.80

ulhu

-0.74

 EntityItem

-0.65

 Vessel

-0.61

DragonMagazine

-0.59

 Norris

-0.59

 Borders

-0.58

 Cotton

-0.56

ħĭ

-0.56

 thumbs

-0.56

POSITIVE LOGITS

based

1.18

sized

1.16

induced

1.00

related

1.00

themed

1.00

level

1.00

derived

0.98

powered

0.98

shaped

0.97

driven

0.95

Activations Density 0.405%

negative opinions or criticism

dash-separated phrases related to various topics

No Comments

No Known Activations