INDEX

Explanations

specific references to religious texts or figures

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

translations and versions

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

 MainAxisSize

-0.70

 للاسماء

-0.66

fromnode

-0.61

+#+#

-0.61

ніципалі

-0.59

Clik

-0.57

WebServlet

-0.57

ſſung

-0.57

 समीक्षाओं

-0.56

 councillors

-0.56

POSITIVE LOGITS

 version

0.53

 versions

0.48

version

0.44

 versões

0.41

 versie

0.40

 Fassung

0.39

 versiones

0.36

 Version

0.36

GIH

0.36

 přek

0.35

Activations Density 0.004%

specific references to religious texts or figures

translations and versions

No Comments

No Known Activations

specific references to religious texts or figures

translations and versions

No Comments

No Known Activations