INDEX

Explanations

objects and surfaces where stickers or paint are applied

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the presence of items attached or affixed to surfaces or locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

uren

-0.77

Â¯Â¯Â¯Â¯

-0.74

================================

-0.68

ORTS

-0.65

cule

-0.65

osponsors

-0.64

ItemThumbnailImage

-0.64

orum

-0.62

================================================================

-0.61

Bu

-0.61

POSITIVE LOGITS

 each

1.16

 every

0.93

top

0.89

each

0.85

 roofs

0.83

the

0.83

top

0.82

 everything

0.80

 behalf

0.78

his

0.77

Activations Density 0.274%

objects and surfaces where stickers or paint are applied

phrases that indicate the presence of items attached or affixed to surfaces or locations

No Comments

No Known Activations