INDEX

Explanations

elements related to data reading and writing structures or operations

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

sports championships

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

 metodología

-0.36

 thème

-0.34

 Gutiérrez

-0.32

$__

-0.31

 conducteur

-0.31

 BoxFit

-0.30

 mécanisme

-0.30

Tikang

-0.30

ViewImports

-0.30

utives

-0.30

POSITIVE LOGITS

 again

0.76

Again

0.73

again

0.71

 Again

0.66

 AGAIN

0.66

 utafitiHapana

0.64

 novamente

0.62

Personendaten

0.57

 autorytatywna

0.57

 <<<<<<<<<<<<<<

0.56

Activations Density 0.285%

elements related to data reading and writing structures or operations

sports championships

No Comments

No Known Activations