INDEX

Explanations

references to mazes or labyrinths

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

maze and labyrinth structures

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ⓘ

-0.41

Officer

-0.39

assertAll

-0.38

 Britton

-0.38

bibitem

-0.38

officer

-0.37

Titles

-0.36

 والث

-0.36

charCodeAt

-0.35

Major

-0.35

POSITIVE LOGITS

 maze

1.35

 Maze

1.29

maze

1.09

Maze

1.04

 labyrinth

1.04

 Labyrinth

0.97

abyrinth

0.95

 laby

0.91

BagLayout

0.61

 Tunnel

0.61

Activations Density 0.004%

references to mazes or labyrinths

maze and labyrinth structures

No Comments

No Known Activations