INDEX

Explanations

versioning information related to software or assemblies

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

start_of_turn user

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

RegressionTest

-0.42

Disclosures

-0.41

mitives

-0.41

 تق

-0.40

DoubleQuotes

-0.40

 Maldonado

-0.39

 meja

-0.39

 trembled

-0.38

govina

-0.37

UnusedPrivate

-0.36

POSITIVE LOGITS

 kasarigan

0.85

 nahilalakip

0.63

 AssemblyVersion

0.62

 AssemblyCulture

0.60

 صوتيه

0.59

verwijspagina

0.54

 defaultstate

0.54

Versión

0.51

 незавершена

0.49

ngdoc

0.49

Activations Density 0.000%

versioning information related to software or assemblies

start_of_turn user

No Comments

No Known Activations