INDEX

Explanations

numerical values or data points mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

numerical data related to legal and administrative contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sonian

-0.69

cffffcc

-0.65

 impat

-0.65

elight

-0.63

rison

-0.62

powers

-0.57

gee

-0.57

especially

-0.57

inventoryQuantity

-0.56

 mentality

-0.56

POSITIVE LOGITS

 apiece

1.22

 total

1.15

 worldwide

1.11

 totaling

1.10

 nationwide

1.04

 spanning

1.04

 globally

1.01

 annually

0.98

per

0.94

 registered

0.93

Activations Density 0.304%

numerical values or data points mentioned in the text

numerical data related to legal and administrative contexts

No Comments

No Known Activations