INDEX

Explanations

sequences of whitespace or line breaks in code or text

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

programming constructs and 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

:✨

-0.80

GraphicsUnit

-0.71

bootstrapcdn

-0.68

 pinulongan

-0.67

 protoimpl

-0.66

цездатний

-0.66

 Chwiliwch

-0.66

 MainAxisSize

-0.65

 NSCoder

-0.64

parsedMessage

-0.64

POSITIVE LOGITS

0.42

 async

0.41

<?

0.41

 simply

0.38

 point

0.37

şti

0.36

ValueGenerated

0.36

 adopt

0.36

教你

0.35

we

0.34

Activations Density 0.001%

sequences of whitespace or line breaks in code or text

programming constructs and symbols

No Comments

No Known Activations