INDEX

Explanations

references to researchers and authors in academic or scientific contexts

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

authors reporting findings

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

Diweddarwch

-0.52

 незавершена

-0.52

RTDA

-0.50

Читати

-0.45

Geplaatst

-0.44

 ujednoznacz

-0.44

 препратки

-0.41

المناصب

-0.41

 createContext

-0.40

 noDo

-0.39

POSITIVE LOGITS

 Lordships

0.57

三人

0.49

LEncoder

0.48

二人

0.47

 Mereka

0.45

HtmlAttribute

0.45

 trebui

0.44

andReturn

0.44

 disponibilités

0.44

四人

0.41

Activations Density 0.038%

references to researchers and authors in academic or scientific contexts

authors reporting findings

No Comments

No Known Activations