INDEX

Explanations

instances of the word "new" in a context suggesting innovation or change

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

-0.62

ebin

-0.60

>>\

-0.60

ouch

-0.59

Ultimately

-0.59

"],"

-0.59

Lau

-0.59

Gamer

-0.58

apor

-0.57

Bra

-0.54

POSITIVE LOGITS

new

2.15

new

2.06

old

1.79

 newer

1.73

NEW

1.70

 newest

1.58

 newly

1.55

New

1.53

old

1.52

NEW

1.52

Activations Density 0.637%

instances of the word "new" in a context suggesting innovation or change

No Comments

No Known Activations