INDEX

Explanations

references to datasets and their availability online

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

links to data files

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

 medarbe

-0.38

 fijo

-0.33

 krę

-0.33

 felicidad

-0.33

 margen

-0.33

 afectada

-0.32

 emitir

-0.32

 ekor

-0.32

 thụ

-0.32

 fjer

-0.32

POSITIVE LOGITS

TypedDataSet

0.66

 disambiguazione

0.65

 дописавши

0.55

HostException

0.52

 zipped

0.52

 downloaded

0.52

 Administrativna

0.51

MessageOf

0.50

 downloading

0.49

elemField

0.48

Activations Density 0.044%

references to datasets and their availability online

links to data files

No Comments

No Known Activations