INDEX

Explanations

phrases related to physical size

oai_token-act-pair · gpt-3.5-turbo

references to size

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

veyard

-0.86

perty

-0.85

roit

-0.83

rique

-0.80

rea

-0.80

INA

-0.79

endez

-0.78

ILLE

-0.78

vous

-0.77

roma

-0.76

POSITIVE LOGITS

 scale

1.04

 stature

0.91

 scales

0.89

 size

0.79

 proportion

0.78

 proportions

0.77

 circumference

0.77

 intestine

0.77

 difference

0.76

 footprint

0.76

Activations Density 0.035%

phrases related to physical size

references to size

No Comments

No Known Activations