INDEX

Explanations

phrases involving possession or association with specific entities or groups

oai_token-act-pair · gpt-3.5-turbo

the word "respective" and its variations, indicating a focus on relationships or comparisons among multiple items or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

hin

-0.64

sted

-0.62

Exception

-0.61

verb

-0.60

iny

-0.56

ek

-0.55

stri

-0.55

won

-0.55

etz

-0.54

zza

-0.53

POSITIVE LOGITS

 respective

3.67

 respectively

1.92

 corresponding

1.77

respective

1.59

 aforementioned

1.31

 afore

1.11

 various

1.09

 counterparts

1.03

 requisite

1.02

selves

0.98

Activations Density 0.005%

phrases involving possession or association with specific entities or groups

the word "respective" and its variations, indicating a focus on relationships or comparisons among multiple items or subjects

No Comments

No Known Activations