INDEX

Explanations

mentions of clothing or costumes

oai_token-act-pair · gpt-3.5-turbo

mentions of clothing or outfits

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

birth

-0.76

elaide

-0.64

Maw

-0.62

ouston

-0.61

asper

-0.61

————————————————

-0.59

Events

-0.59

 Nanto

-0.59

Lenin

-0.58

osate

-0.58

POSITIVE LOGITS

ters

1.00

 worn

0.89

ments

0.89

 apparel

0.87

swick

0.86

 outfits

0.81

ting

0.79

glers

0.79

uated

0.77

 outfit

0.74

Activations Density 0.016%

mentions of clothing or costumes

mentions of clothing or outfits

No Comments

No Known Activations