INDEX

Explanations

references to social media platforms, particularly Facebook and Twitter

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

social media platforms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-0.81

DrawerToggle

-0.59

MessageOf

-0.55

MessageTagHelper

-0.51

 disambiguazione

-0.50

InstrumentedTest

-0.50

 utafitiHapana

-0.50

tagHelperRunner

-0.48

UrlResolution

-0.47

 invokingState

-0.46

POSITIVE LOGITS

 Facebook

0.42

 LinkedIn

0.40

 platform

0.39

 FACEBOOK

0.37

 marketplace

0.36

 facebook

0.35

LinkedIn

0.34

lda

0.33

Facebook

0.33

TWITTER

0.33

Activations Density 0.028%

references to social media platforms, particularly Facebook and Twitter

social media platforms

No Comments

No Known Activations