INDEX

Explanations

phrases related to hidden or valuable items

oai_token-act-pair · gpt-3.5-turbo

references to valuable items or concepts metaphorically described as "treasures."

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

LINE

-0.72

SPA

-0.70

sense

-0.68

ells

-0.68

okin

-0.66

ellar

-0.66

imov

-0.66

66666666

-0.65

enegger

-0.65

abad

-0.63

POSITIVE LOGITS

 trove

1.26

abilia

1.05

 treasure

0.98

 Treasure

0.92

 treasures

0.90

asures

0.81

 chests

0.79

 riches

0.77

asure

0.74

ngth

0.73

Activations Density 0.009%

phrases related to hidden or valuable items

references to valuable items or concepts metaphorically described as "treasures."

No Comments

No Known Activations