INDEX

Explanations

phrases related to the concept of "having a hole in something."

oai_token-act-pair · gpt-3.5-turbo

references to physical openings or gaps 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

APH

-0.82

wcsstore

-0.79

Ô

-0.79

umers

-0.72

zens

-0.70

esses

-0.70

onym

-0.70

ribes

-0.68

imei

-0.67

assadors

-0.67

POSITIVE LOGITS

 pavement

0.83

 fabric

0.79

 ceiling

0.78

 fence

0.77

sec

0.77

 membranes

0.73

 coverage

0.70

 limestone

0.70

 walls

0.69

 hull

0.68

Activations Density 0.226%

phrases related to the concept of "having a hole in something."

references to physical openings or gaps in various contexts

No Comments

No Known Activations