INDEX

Explanations

words related to a specific single word, "monolith."

oai_token-act-pair · gpt-3.5-turbo

instances of the prefix "mon," suggesting a focus on words related to singularity or dominance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

INGTON

-0.83

 Norn

-0.80

FUL

-0.73

ORTS

-0.70

PORT

-0.69

PLAY

-0.68

 Venom

-0.67

UGE

-0.66

EMP

-0.65

ACTED

-0.65

POSITIVE LOGITS

stros

1.52

olithic

1.44

olith

1.34

ograph

1.30

ocular

1.25

oton

1.24

ographs

1.23

ologue

1.20

soon

1.17

ospace

1.16

Activations Density 0.023%

words related to a specific single word, "monolith."

instances of the prefix "mon," suggesting a focus on words related to singularity or dominance

No Comments

No Known Activations