INDEX

Explanations

keywords related to functions and features of products or systems

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

installation and applications

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

本

-0.38

"/",

-0.34

%@",

-0.34

Dijo

-0.33

 HasFactory

-0.32

Становништво

-0.31

 hech

-0.30

 herv

-0.30

 Schwan

-0.29

 Dijo

-0.29

POSITIVE LOGITS

inWeight

0.52

 installations

0.52

Installation

0.51

glove

0.51

 installation

0.50

 Machinery

0.50

ỡng

0.49

Personendaten

0.49

 disambiguazione

0.49

 OSHA

0.48

Activations Density 0.018%

keywords related to functions and features of products or systems

installation and applications

No Comments

No Known Activations