INDEX

Explanations

references to user review platforms and associated ratings

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

Yelp, Google, reviews, ratings

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

resave

-0.35

Pautan

-0.33

<?

-0.32

恿

-0.31

それで

-0.31

 kaarangay

-0.31

subpackage

-0.30

 pulau

-0.30

 desperation

-0.29

Asts

-0.29

POSITIVE LOGITS

VYMaps

0.79

 Yelp

0.72

haikusbot

0.60

 GOOGLE

0.57

 Google

0.57

UPAC

0.56

oursquare

0.56

 google

0.56

Google

0.55

GOOGLE

0.54

Activations Density 0.001%

references to user review platforms and associated ratings

Yelp, Google, reviews, ratings

No Comments

No Known Activations