INDEX

Explanations

sentences typically found in formal written correspondence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating uncertain or vague statements

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

 sqor

-0.80

oun

-0.71

sembly

-0.68

 commissions

-0.67

 metic

-0.67

 occas

-0.66

 streng

-0.64

 obser

-0.64

 princ

-0.64

 indo

-0.63

POSITIVE LOGITS

1.64

"[

1.61

"'

1.57

"(

1.54

'"

1.27

"...

1.26

"âĢ¦

1.24

Shape

1.20

SPONSORED

1.16

Advertisement

1.14

Activations Density 0.122%

sentences typically found in formal written correspondence

phrases indicating uncertain or vague statements

No Comments

No Known Activations