INDEX

Explanations

statements from spokespersons or representatives

oai_token-act-pair · gpt-3.5-turbo

Instances of the word "said"

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

otin

-0.89

à¦

-0.84

=~=~

-0.77

ï¸

-0.69

advertising

-0.69

 ILCS

-0.67

otion

-0.67

COR

-0.66

Holy

-0.66

âĸ

-0.65

POSITIVE LOGITS

doms

0.94

 bluntly

0.79

 anecd

0.79

 afterward

0.77

 afterwards

0.73

 goodbye

0.71

 sarcast

0.69

 yesterday

0.67

 earlier

0.65

heit

0.64

Activations Density 0.307%

statements from spokespersons or representatives

Instances of the word "said"

No Comments

No Known Activations