INDEX

Explanations

formatted strings and placeholders in code snippets

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

Python f-strings and format specifiers

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

RegressionTest

-0.71

TargetException

-0.68

enumii

-0.60

PreferredItem

-0.58

 EconPapers

-0.56

HideFlags

-0.56

__':

-0.56

+#+#

-0.54

ChildScrollView

-0.54

 BoxFit

-0.54

POSITIVE LOGITS

 uska

0.42

 Krieg

0.37

docx

0.36

 Trock

0.36

 här

0.35

ilität

0.35

einander

0.34

Carreira

0.34

milch

0.33

 Española

0.33

Activations Density 0.016%

formatted strings and placeholders in code snippets

Python f-strings and format specifiers

No Comments

No Known Activations