INDEX

Explanations

numerical data and references to time or duration

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

respectively, inclusive, demonstrating

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

GenerationType

-0.62

+#+#

-0.56

 &___

-0.49

 AssemblyCompany

-0.47

 defaultstate

-0.46

KommentareTeilen

-0.46

GOTREF

-0.45

liferay

-0.45

 informée

-0.45

 leaſt

-0.44

POSITIVE LOGITS

 betrek

0.49

новништво

0.49

 przeznac

0.47

 featuring

0.45

 jotka

0.45

 betreft

0.45

⟬

0.45

 obsah

0.44

 containing

0.44

 whose

0.44

Activations Density 0.046%

numerical data and references to time or duration

respectively, inclusive, demonstrating

No Comments

No Known Activations