INDEX

Explanations

statements or clarifications regarding events or actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tagHelperRunner

-0.73

تقاوى

-0.44

 Roskov

-0.39

 moving

-0.33

onViewCreated

-0.32

DockStyle

-0.31

 defaultstate

-0.30

UNI

-0.30

 mait

-0.30

uni

-0.28

POSITIVE LOGITS

ArgsConstructor

0.57

 apologized

0.55

archiwizowane

0.55

softc

0.53

 AssemblyTitle

0.53

 apologised

0.53

Rüyada

0.52

RegressionTest

0.51

 lendemain

0.50

つもり

0.50

Activations Density 0.062%

statements or clarifications regarding events or actions

No Comments

No Known Activations