INDEX

Explanations

programming syntax and structure

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

opt out from contacts

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

!*\

-0.54

WriteTagHelper

-0.52

 resourceCulture

-0.48

UrlResolution

-0.46

httphttps

-0.45

ronic

-0.45

matchCondition

-0.45

medriver

-0.44

angliski

-0.43

 TestBed

-0.42

POSITIVE LOGITS

 TextAppearance

0.49

+:+

0.46

 MenuView

0.46

Jîn

0.44

MessageWindow

0.43

 kasarigan

0.42

 informée

0.41

avadoc

0.40

Kjelder

0.40

 referenties

0.39

Activations Density 0.034%

programming syntax and structure

opt out from contacts

No Comments

No Known Activations