INDEX

Explanations

programming concepts and terminology related to classes and object-oriented design

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

Batting position

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

 suaminya

-0.47

InstanceState

-0.45

 voork

-0.42

 dignité

-0.42

 Komunikasi

-0.41

bentar

-0.41

 pazi

-0.41

FieldOffsetTable

-0.41

 loopt

-0.41

 būs

-0.41

POSITIVE LOGITS

verwijspagina

0.58

 Wikimedijinoj

0.51

FormTagHelper

0.50

ControllerAdvice

0.49

 propOrder

0.47

Aktualisiert

0.47

ValueStyle

0.46

Derbyniad

0.46

CURLOPT

0.46

AnchorStyles

0.45

Activations Density 0.021%

programming concepts and terminology related to classes and object-oriented design

Batting position

No Comments

No Known Activations