INDEX

Explanations

the word "able" followed by a number, likely indicating capability or possibility

oai_token-act-pair · gpt-3.5-turbo

phrases indicating capability or ability

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

bane

-0.70

 Kendall

-0.67

 Tradition

-0.64

 Clemson

-0.63

origin

-0.63

 Caval

-0.62

 Origin

-0.62

 Warriors

-0.62

 Traffic

-0.61

 Belief

-0.60

POSITIVE LOGITS

bodied

0.98

anooga

0.93

ioned

0.87

reys

0.84

osaurus

0.83

¶æ

0.83

't

0.83

 inval

0.81

iculty

0.79

awaru

0.78

Activations Density 0.028%

the word "able" followed by a number, likely indicating capability or possibility

phrases indicating capability or ability

No Comments

No Known Activations