INDEX

Explanations

positive and negative personality traits

oai_token-act-pair · gpt-3.5-turbo

descriptors related to strong moral principles and values

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

weeney

-0.75

swick

-0.70

abet

-0.69

ModLoader

-0.67

adelphia

-0.66

and

-0.66

rity

-0.64

or

-0.64

et

-0.62

adobe

-0.61

POSITIVE LOGITS

 alike

1.61

 respectively

1.02

 depending

0.76

depending

0.73

 ways

0.68

 environments

0.66

 sounding

0.62

 manner

0.61

 enough

0.60

 avenues

0.59

Activations Density 0.292%

positive and negative personality traits

descriptors related to strong moral principles and values

No Comments

No Known Activations