INDEX

Explanations

instances of variable assignments or definitions in coding contexts

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

scale ranges

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

IntoConstraints

-0.65

SupportActionBar

-0.56

tagHelperRunner

-0.52

Бахар

-0.51

xes

-0.46

 betweenstory

-0.46

__*/

-0.46

 peas

-0.46

StreetMap

-0.46

erialized

-0.45

POSITIVE LOGITS

 الرياضيه

0.48

 "..\..\

0.39

 "..\..\..\

0.38

цезда

0.34

🟤

0.33

embolso

0.33

 Numerade

0.33

siempre

0.32

tonode

0.31

may

0.31

Activations Density 0.063%

instances of variable assignments or definitions in coding contexts

scale ranges

No Comments

No Known Activations