INDEX

Explanations

mentions of the term "Kin" with varying numerical associations

oai_token-act-pair · gpt-3.5-turbo

variations and references to the concept of "kin" or relationships among family and close individuals

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

enance

-0.90

shire

-0.65

Ö¼

-0.63

 Seym

-0.62

 Mellon

-0.62

channelAvailability

-0.61

 loopholes

-0.59

 Ronaldo

-0.58

 fine

-0.58

 Verd

-0.58

POSITIVE LOGITS

etics

1.49

etic

1.41

etically

1.32

folk

1.14

sey

1.13

esthetic

1.10

ners

1.07

ja

1.05

ned

1.05

zie

1.03

Activations Density 0.039%

mentions of the term "Kin" with varying numerical associations

variations and references to the concept of "kin" or relationships among family and close individuals

No Comments

No Known Activations