INDEX

Explanations

terms related to processes and actions in various contexts

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

everything + description

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

werde

-0.45

 sprüche

-0.44

 ansatte

-0.43

odis

-0.40

 plantilla

-0.40

 uitz

-0.40

ñera

-0.39

ulipas

-0.39

 blanda

-0.39

 Mujeres

-0.38

POSITIVE LOGITS

everything

1.00

 everything

0.99

 Everything

0.98

Everything

0.95

 EVERYTHING

0.85

 tudo

0.71

的一切

0.66

一切都

0.63

 wszystko

0.62

 alles

0.60

Activations Density 0.040%

terms related to processes and actions in various contexts

everything + description

No Comments

No Known Activations