INDEX

Explanations

markers that indicate various data states or conditions in a procedural context

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

technical terms and foreign phrases

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

jspx

-0.46

etag

-0.40

 nonatomic

-0.39

Décès

-0.38

новништво

-0.37

TW

-0.37

alisa

-0.37

 defeat

-0.36

ENESS

-0.36

Jef

-0.35

POSITIVE LOGITS

Rhestr

0.73

IntoConstraints

0.46

devamını

0.45

 ویکی‌آمباردا

0.45

 Normdatei

0.44

 Italijanski

0.44

bootstrapcdn

0.43

RTSN

0.43

phazard

0.42

ModelSerializer

0.42

Activations Density 0.183%

markers that indicate various data states or conditions in a procedural context

technical terms and foreign phrases

No Comments

No Known Activations