INDEX

Explanations

references to political oppression and authoritarian regimes

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

groups associated with violence and occupation

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

 halo

-0.46

 Leclerc

-0.45

 crown

-0.44

MENAFN

-0.43

 Halo

-0.43

RTSC

-0.43

олові

-0.42

//

-0.41

 Crown

-0.41

Halo

-0.41

POSITIVE LOGITS

 nahilalakip

0.49

+};

0.47

UnusedPrivate

0.42

 FetchType

0.41

 ويكيميديا

0.35

 استنادى

0.35

>>()

0.35

 bodyParser

0.35

 ब्रेकडाउन

0.35

"}\

0.34

Activations Density 0.037%

references to political oppression and authoritarian regimes

groups associated with violence and occupation

No Comments

No Known Activations