INDEX

Explanations

instances of mathematical notation and formatting elements

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

labels and code blocks

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ĝis

-0.45

ilight

-0.43

neurial

-0.43

licit

-0.43

どり

-0.41

참고

-0.41

ikki

-0.40

tadiene

-0.40

McN

-0.40

 Afro

-0.40

POSITIVE LOGITS

yntaxException

0.47

 surla

0.45

 للاسماء

0.44

 disponibilités

0.43

transQ

0.43

########.

0.41

StoryboardSegue

0.39

 propOrder

0.39

 AssemblyCompany

0.39

 geométrico

0.38

Activations Density 0.201%

instances of mathematical notation and formatting elements

labels and code blocks

No Comments

No Known Activations