INDEX

Explanations

references to alternative entities or subjects

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

other followed by noun

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

XmlAccessType

-0.47

MethodImpl

-0.46

Portail

-0.45

 Shakspeare

-0.44

 Nehru

-0.44

 Selangor

-0.43

Geplaatst

-0.43

Kuj

-0.41

onauts

-0.41

*)&

-0.41

POSITIVE LOGITS

 OTHER

0.60

もう

0.58

 other

0.56

enumi

0.55

MLLoader

0.54

 Other

0.53

 остальных

0.53

 another

0.52

 ANOTHER

0.52

 TAMBÉM

0.52

Activations Density 0.019%

references to alternative entities or subjects

other followed by noun

No Comments

No Known Activations

references to alternative entities or subjects

other followed by noun

No Comments

No Known Activations