INDEX

Explanations

references to statistical data and experimental results

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

citations, lists, and technical terms

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

rungsseite

-0.46

Билгалдахарш

-0.44

 surla

-0.42

примеча

-0.40

Espèce

-0.38

########.

-0.36

tilhe

-0.36

tellation

-0.36

')";

-0.36

قایناق‌لار

-0.36

POSITIVE LOGITS

 resourceCulture

0.46

 يتيمه

0.41

msgTypes

0.40

delwed

0.40

clés

0.39

 coefficient

0.39

Eigen

0.38

 kalan

0.38

'{@

0.38

 Bezir

0.37

Activations Density 0.130%

references to statistical data and experimental results

citations, lists, and technical terms

No Comments

No Known Activations