INDEX

Explanations

references to specific software licenses and their terms

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

Armonk copyright license

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

 apapun

-0.44

 yoktur

-0.43

 Behörden

-0.42

 prohibido

-0.42

transQ

-0.41

 Quellen

-0.40

 Betreuung

-0.40

 Verteilung

-0.40

 Schmid

-0.40

 Decke

-0.40

POSITIVE LOGITS

ésultats

0.61

ðsíða

0.59

 Roskov

0.58

arangay

0.56

 bandou

0.56

gof

0.52

iſen

0.52

éphane

0.52

 blin

0.50

hematical

0.50

Activations Density 0.031%

references to specific software licenses and their terms

Armonk copyright license

No Comments

No Known Activations