INDEX

Explanations

references and elements related to mathematical notation and figures

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

fig references with labels

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

IInterface

-0.26

 refusé

-0.26

 party

-0.25

vok

-0.25

 inférieur

-0.25

Initializable

-0.25

party

-0.25

 قر

-0.25

!*\

-0.24

-0.23

POSITIVE LOGITS

 nakalista

0.83

rungsseite

0.75

 <<<<<<<<<<<<<<

0.73

verifyException

0.70

setVerticalGroup

0.68

 للاسماء

0.67

 NSCoder

0.64

énario

0.63

 AssemblyProduct

0.62

 snippetHide

0.60

Activations Density 0.026%

references and elements related to mathematical notation and figures

fig references with labels

No Comments

No Known Activations