INDEX

Explanations

references to scientific figures and tables

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

PLOS ONE article identifiers

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

 Neves

-0.32

 system

-0.32

 environment

-0.32

пущен

-0.31

__;

-0.31

Vege

-0.30

 within

-0.29

 area

-0.29

Учитель

-0.28

POSITIVE LOGITS

دانشنامهٔ

0.71

出版年

0.66

NameInMap

0.65

 nahilalakip

0.60

VersionUID

0.60

 Numerade

0.60

 ویکی‌پدیا

0.59

RUnlock

0.58

 autorytatywna

0.57

Chham

0.57

Activations Density 0.918%

references to scientific figures and tables

PLOS ONE article identifiers

No Comments

No Known Activations