INDEX

Explanations

references to scientific entities and their properties

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

describing media and tools

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

 geç

-0.30

<eos>

-0.29

↵

-0.26

č

-0.26

is

-0.25

 demais

-0.25

 arcilla

-0.25

POSITIVE LOGITS

DockStyle

0.94

 disambiguazione

0.81

 initComponents

0.79

tvguidetime

0.76

 AssemblyCulture

0.71

rungsseite

0.71

出版年

0.71

 للاسماء

0.70

NUMX

0.70

 utafitiHapana

0.70

Activations Density 0.246%

references to scientific entities and their properties

describing media and tools

No Comments

No Known Activations