INDEX

Explanations

phrases related to previous expectations and beliefs

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

comparison to expectation

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

writerow

-0.52

 elettrica

-0.42

 käs

-0.42

GIVEREF

-0.41

 omge

-0.41

{}",

-0.40

 pierna

-0.40

 ympä

-0.39

Rhestr

-0.38

 egli

-0.38

POSITIVE LOGITS

ModelAdmin

0.60

 ujednoznacz

0.57

ViewFeatures

0.53

 autorytatywna

0.50

 invokingState

0.48

ModelSerializer

0.42

 Nina

0.41

قایناقلار

0.41

Amelia

0.41

Посилання

0.41

Activations Density 0.014%

phrases related to previous expectations and beliefs

comparison to expectation

No Comments

No Known Activations