INDEX

Explanations

personal opinions or evaluations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "it" in various contexts

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

 ILCS

-0.65

odder

-0.62

ãĥ¯ãĥ³

-0.62

èĢħ

-0.61

è¦ļéĨĴ

-0.60

 è£ıè¦ļéĨĴ

-0.60

estern

-0.59

shoot

-0.59

pring

-0.57

cribed

-0.56

POSITIVE LOGITS

 nonetheless

1.28

 does

1.24

 nevertheless

1.23

 doesn

1.22

isn

1.19

 DOES

1.19

ain

1.18

 certainly

1.13

 shouldn

1.13

 hasn

1.12

Activations Density 0.133%

personal opinions or evaluations

instances of the word "it" in various contexts

No Comments

No Known Activations