INDEX

Explanations

phrases related to large numerical measurements

oai_token-act-pair · gpt-3.5-turbo

references to measurements, particularly related to dimensions and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 moderation

-0.66

 sides

-0.66

 Creator

-0.66

 Scrib

-0.65

 Eternity

-0.64

 fundament

-0.64

 Twist

-0.64

 Rica

-0.63

 robbing

-0.63

 Clifford

-0.62

POSITIVE LOGITS

pound

1.36

plus

1.30

capacity

1.13

page

1.13

strong

1.11

mile

1.08

degree

1.08

dollar

1.06

year

1.06

odd

1.05

Activations Density 0.063%

phrases related to large numerical measurements

references to measurements, particularly related to dimensions and quantities

No Comments

No Known Activations