INDEX

Explanations

programming and scripting constructs or statements

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

numerical quantities

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 AssemblyCulture

-1.05

complexContent

-0.73

WriteTagHelper

-0.69

 transfieras

-0.68

DockStyle

-0.68

::_('

-0.67

 jLabel

-0.66

__::

-0.65

UserScript

-0.65

存于互联网档案馆

-0.64

POSITIVE LOGITS

 dozen

0.55

 hundred

0.47

 thousand

0.45

 dozens

0.44

 cuantos

0.42

 dizaine

0.42

 sintético

0.42

 hundert

0.42

 salvajes

0.42

 escolares

0.41

Activations Density 0.058%

programming and scripting constructs or statements

numerical quantities

No Comments

No Known Activations