INDEX

Explanations

phrases mentioning common sense or commonality

oai_token-act-pair · gpt-3.5-turbo

references to "common sense" and its variations in different 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

 Become

-0.74

zona

-0.74

asus

-0.72

endas

-0.70

letes

-0.69

atoon

-0.67

hetamine

-0.66

otos

-0.66

agate

-0.66

awaru

-0.65

POSITIVE LOGITS

wealth

1.63

alities

1.41

ality

1.28

 denomin

1.26

places

1.17

place

1.02

 ancestor

1.02

 sense

1.00

 decency

0.97

alties

0.91

Activations Density 0.034%

phrases mentioning common sense or commonality

references to "common sense" and its variations in different contexts

No Comments

No Known Activations