INDEX

Explanations

phrases that reflect hypothetical scenarios or conditional statements

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

and it

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

webElementXpaths

-0.54

évaluateur

-0.47

Rhestr

-0.47

PasswordField

-0.43

ImageContext

-0.43

🇶

-0.41

 propOrder

-0.41

ViewImports

-0.41

 kaarangay

-0.40

 సౌకర్య

-0.40

POSITIVE LOGITS

WebServlet

0.44

cshtml

0.44

 оригіналу

0.42

 fréqu

0.40

 CURIAM

0.36

IBOutlet

0.36

altezza

0.36

resizingMask

0.34

Ча

0.34

 stanga

0.34

Activations Density 0.028%

phrases that reflect hypothetical scenarios or conditional statements

and it

No Comments

No Known Activations