INDEX

Explanations

programming language syntax and formatting details

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

language and plural forms

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

səhifə

-0.47

 الدولى

-0.45

laubnis

-0.44

ernalia

-0.43

 binocular

-0.43

 Normdatei

-0.42

pecabezas

-0.42

Fourni

-0.42

 NACIONAL

-0.42

rotu

-0.42

POSITIVE LOGITS

########.

0.52

ContentAsync

0.47

IUrlHelper

0.46

posedge

0.44

Merkmale

0.43

AISSEE

0.41

قایناقلار

0.40

 AssemblyCompany

0.38

atisk

0.36

endphp

0.36

Activations Density 0.002%

programming language syntax and formatting details

language and plural forms

No Comments

No Known Activations