INDEX

Explanations

references to mazes or labyrinth-like structures

oai_token-act-pair · gpt-3.5-turbo

terms related to mazes and labyrinths

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

fman

-0.78

imates

-0.74

ople

-0.72

arer

-0.70

zinski

-0.70

izations

-0.69

oples

-0.69

mentation

-0.69

utical

-0.67

ment

-0.67

POSITIVE LOGITS

 maze

1.23

 labyrinth

1.00

yrinth

0.86

ĸļ

0.86

stakes

0.85

 Maze

0.81

 Turing

0.79

 Morse

0.75

 convoluted

0.74

 Parkway

0.74

Activations Density 0.022%

references to mazes or labyrinth-like structures

terms related to mazes and labyrinths

No Comments

No Known Activations