INDEX

Explanations

sequences of whitespace or non-visible characters

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

public getters and setters

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

 betweenstory

-1.14

transQ

-1.11

GEBURTSDATUM

-1.09

 Biôgrafia

-1.01

uxxxx

-1.01

 ब्रेकडाउन

-0.96

Autoritní

-0.95

WriteTagHelper

-0.94

afficheront

-0.93

 Italijanski

-0.91

POSITIVE LOGITS

 either

0.31

 which

0.31

 dobles

0.31

 Reino

0.30

 double

0.28

 Romains

0.28

 nég

0.27

 they

0.27

 Feind

0.27

 Pendidikan

0.26

Activations Density 0.000%

sequences of whitespace or non-visible characters

public getters and setters

No Comments

No Known Activations