INDEX

Explanations

mathematical or programming-related notation and structures

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

principalColumn

-0.62

FunctionFlags

-0.56

 فريبيس

-0.54

__.__

-0.52

ToBounds

-0.52

ftagPool

-0.51

참고

-0.50

 AssemblyCompany

-0.49

onOptions

-0.49

RegistryLite

-0.49

POSITIVE LOGITS

 AppBundle

0.37

 context

0.36

Kli

0.36

 Noire

0.35

cli

0.35

Kir

0.34

Re

0.33

CHOICES

0.33

 Hers

0.33

撚

0.33

Activations Density 0.021%

mathematical or programming-related notation and structures

No Comments

No Known Activations