INDEX

Explanations

terms related to social media usage and engagement

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

social media

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

Negative Logits

DockStyle

-0.49

tagHelperRunner

-0.47

shadowOpacity

-0.45

ContentAlignment

-0.44

تقاوى

-0.42

 Entwurf

-0.41

CGRectMake

-0.39

 hır

-0.38

 pernas

-0.38

TypedDataSet

-0.38

POSITIVE LOGITS

 facebook

0.64

 FACEBOOK

0.63

 Facebook

0.63

 estekak

0.61

 Instagram

0.60

Facebook

0.59

 twitter

0.59

 Twitter

0.59

 instagram

0.59

Instagram

0.57

Activations Density 0.004%

terms related to social media usage and engagement

social media

No Known Activations