INDEX

Explanations

words related to providing examples or offering explanations

oai_token-act-pair · gpt-3.5-turbo

instances of "being" and examples that illustrate concepts or arguments

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rity

-0.66

roup

-0.64

hered

-0.63

 swore

-0.62

elf

-0.60

rosis

-0.60

ahs

-0.60

aired

-0.59

ioned

-0.58

IRE

-0.58

POSITIVE LOGITS

 example

1.64

 examples

1.50

example

1.42

 illustration

1.38

 illust

1.28

 exception

1.22

 exempl

1.22

 Exhibit

1.20

 proof

1.15

 Example

1.10

Activations Density 0.612%

words related to providing examples or offering explanations

instances of "being" and examples that illustrate concepts or arguments

No Comments

No Known Activations