INDEX

Explanations

names of institutions or organizations, particularly in healthcare and education contexts

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

organizations and places

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

NewUrlParser

-0.48

ſelves

-0.47

Personendaten

-0.47

ftagPool

-0.47

 poffe

-0.47

Ɓ

-0.45

 Italijanski

-0.44

 kasarigan

-0.43

 Bulgar

-0.43

 beſt

-0.43

POSITIVE LOGITS

in

0.49

Décès

0.48

 near

0.47

帖最后由

0.47

 located

0.47

GraphicsUnit

0.43

بوابة

0.42

แห่ง

0.42

AnchorStyles

0.41

 North

0.40

Activations Density 0.048%

names of institutions or organizations, particularly in healthcare and education contexts

organizations and places

No Comments

No Known Activations