INDEX

Explanations

references to user-generated content and ratings

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

review dates

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

Tikang

-0.58

تقاوى

-0.46

сылкі

-0.45

 intptr

-0.44

URLException

-0.44

afficheront

-0.44

ChromeDriver

-0.44

subpackage

-0.42

 Мексичка

-0.42

 CreateTagHelper

-0.42

POSITIVE LOGITS

 opérés

0.47

 reviewer

0.42

*/}

0.39

 становника

0.39

 GreatSchools

0.38

Caller

0.37

Affirmed

0.36

Bundle

0.36

lapsingToolbar

0.36

 Ratings

0.36

Activations Density 0.011%

references to user-generated content and ratings

review dates

No Comments

No Known Activations