INDEX

Explanations

terms related to programming syntax and data handling

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

numbers and code symbols

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

 CURIAM

-0.73

 mxArray

-0.70

XmlAccessorType

-0.65

 للاسماء

-0.62

rungsseite

-0.58

PerformLayout

-0.58

endphp

-0.57

 intptr

-0.54

ConstraintMaker

-0.52

 kasarigan

-0.52

POSITIVE LOGITS

0.47

／

0.34

or

0.33

LocalStorage

0.32

unki

0.31

ädie

0.31

 Chriſt

0.30

Fernseh

0.30

 meille

0.30

Activations Density 0.547%

terms related to programming syntax and data handling

numbers and code symbols

No Comments

No Known Activations