INDEX

Explanations

content related to historical events or documents

oai_token-act-pair · gpt-3.5-turbo

references to specific splits and sagas, particularly related to the "Dark Phoenix" story

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

inders

-0.71

 Hess

-0.70

 seism

-0.68

 benz

-0.68

 concess

-0.67

eware

-0.65

èĥ

-0.65

urally

-0.65

 unwelcome

-0.64

oor

-0.63

POSITIVE LOGITS

 Split

3.41

Split

1.79

 splits

1.36

split

1.28

 Saga

1.21

Spl

1.19

acan

1.12

Spl

1.01

uku

1.01

spl

0.94

Activations Density 0.047%

content related to historical events or documents

references to specific splits and sagas, particularly related to the "Dark Phoenix" story

No Comments

No Known Activations