INDEX

Explanations

references to search queries and information retrieval processes

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

Google searches and results

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

 betweenstory

-0.43

лактика

-0.36

 Olli

-0.31

具

-0.31

Geografi

-0.31

 grace

-0.30

 heller

-0.30

 Utama

-0.30

 ultimately

-0.30

 merak

-0.30

POSITIVE LOGITS

 resourceCulture

0.62

errHandler

0.60

脚注の使い方

0.59

afficheront

0.56

VIAF

0.52

IntoConstraints

0.50

 Walkover

0.49

rinfo

0.48

 Wiktionnaire

0.47

TypedDataSet

0.47

Activations Density 0.105%

references to search queries and information retrieval processes

Google searches and results

No Comments

No Known Activations