INDEX

Explanations

references to gaining something or acquiring benefits

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "gain" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

pled

-0.76

phis

-0.63

 reckoned

-0.60

 contemplated

-0.59

 densely

-0.59

 Mole

-0.58

 Dalai

-0.57

 Seriously

-0.57

Pol

-0.56

 Gears

-0.56

POSITIVE LOGITS

esville

1.05

 gain

0.96

llor

0.85

 gains

0.85

stadt

0.81

icult

0.80

/+

0.80

itures

0.79

esses

0.79

iture

0.78

Activations Density 0.010%

references to gaining something or acquiring benefits

references to the concept of "gain" in various contexts

No Comments

No Known Activations