INDEX

Explanations

phrases describing extreme or superlative experiences

oai_token-act-pair · gpt-3.5-turbo

expressions of extreme positive or negative experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

gravity

-0.67

 Cosponsors

-0.62

phasis

-0.59

 basics

-0.57

Phi

-0.54

urat

-0.54

rocal

-0.54

 predomin

-0.54

çīĪ

-0.54

uta

-0.53

POSITIVE LOGITS

 ever

1.62

 EVER

1.38

 Ever

1.24

ever

1.21

Ever

1.18

 imaginable

1.17

've

0.95

 possibly

0.90

 could

0.82

 conceivable

0.81

Activations Density 0.112%

phrases describing extreme or superlative experiences

expressions of extreme positive or negative experiences

No Comments

No Known Activations