INDEX

Explanations

phrases related to attaching or mounting objects to other objects

oai_token-act-pair · gpt-3.5-turbo

phrases related to attachments and mounting objects

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

osponsors

-0.79

displayText

-0.74

Reports

-0.70

MQ

-0.69

IQ

-0.68

Ô

-0.67

ivas

-0.67

ild

-0.66

ItemThumbnailImage

-0.66

quickShipAvailable

-0.65

POSITIVE LOGITS

 underside

0.82

top

0.77

 holes

0.77

 walls

0.70

 bushes

0.69

 racks

0.69

estone

0.69

 concrete

0.69

 tissue

0.68

 each

0.68

Activations Density 0.200%

phrases related to attaching or mounting objects to other objects

phrases related to attachments and mounting objects

No Comments

No Known Activations