INDEX

Explanations

keywords and phrases related to technical specifications and details

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

described using following words

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

ścio

-0.39

ilon

-0.38

<bos>

-0.37

FirstName

-0.35

abase

-0.34

ABASE

-0.34

 fréqu

-0.34

bkz

-0.34

 Bowling

-0.34

har

-0.33

POSITIVE LOGITS

parsedMessage

0.66

featureID

0.64

SharedDtor

0.62

providedIn

0.60

 Taktlose

0.58

ViewInit

0.57

 pleaſure

0.54

InjectAttribute

0.54

addCriterion

0.53

الحياه

0.52

Activations Density 0.132%

keywords and phrases related to technical specifications and details

described using following words

No Comments

No Known Activations