INDEX

Explanations

references to numerical values and data points in various contexts

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

conductive layer, kinase, tire, baby

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

imp

-0.29

 SHALL

-0.28

 quality

-0.28

 shall

-0.26

som

-0.25

etc

-0.25

↵

-0.24

ty

-0.24

jem

-0.23

 mark

-0.23

POSITIVE LOGITS

ChildScrollView

0.88

 ujednoznacz

0.77

parsedMessage

0.72

CloseOperation

0.72

 snippetHide

0.71

 '\\;'

0.71

 createSprite

0.70

 دیکھیے

0.70

bewerken

0.68

 Catawiki

0.65

Activations Density 0.018%

references to numerical values and data points in various contexts

conductive layer, kinase, tire, baby

No Comments

No Known Activations