INDEX

Explanations

mentions of the word "Sen" followed by a number

oai_token-act-pair · gpt-3.5-turbo

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dawn

-0.73

Reference

-0.64

 Heard

-0.64

selves

-0.63

ATHER

-0.63

 Carmen

-0.63

 Cind

-0.62

 Bearing

-0.61

ORN

-0.60

 Xander

-0.59

POSITIVE LOGITS

iors

1.52

egal

1.49

pai

1.47

seless

1.39

eca

1.31

escent

1.31

hov

1.10

ryu

1.07

esc

1.02

esis

0.99

Activations Density 0.030%

mentions of the word "Sen" followed by a number

No Comments

No Known Activations