INDEX

Explanations

comparative phrases using "more for less."

oai_token-act-pair · gpt-3.5-turbo

negative descriptors or criticism in a context that may imply moral judgment

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Neb

-0.55

 Petraeus

-0.53

 IMAGES

-0.52

Picture

-0.51

 Axel

-0.50

NBN

-0.49

Pog

-0.49

 Fallout

-0.48

 Niet

-0.48

Nap

-0.46

POSITIVE LOGITS

 respectively

0.97

}.

0.76

]).

0.72

]."

0.69

anwhile

0.65

.).

0.63

)).

0.62

äº

0.61

lambda

0.61

*.

0.60

Activations Density 6.024%

comparative phrases using "more for less."

negative descriptors or criticism in a context that may imply moral judgment

No Comments

No Known Activations