INDEX

Explanations

code-related syntax, particularly semicolons and declarations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 Commencez

-0.70

 domestiques

-0.56

wijl

-0.54

 Veuillez

-0.52

 domestique

-0.50

 quelqu

-0.49

 devons

-0.49

entibus

-0.48

 Porphy

-0.48

 costumi

-0.47

POSITIVE LOGITS

SharedDtor

0.74

InstrumentedTest

0.71

CppMethod

0.68

prepareStatement

0.68

 дописавши

0.66

elemField

0.63

 protoimpl

0.63

tagHelperRunner

0.62

astify

0.61

setArguments

0.61

Activations Density 0.183%

code-related syntax, particularly semicolons and declarations

No Comments

No Known Activations