INDEX

Explanations

words and phrases related to decision-making and processes involving claims and reviews

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

will send email

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

 numberOfRows

-0.42

lauf

-0.40

 appunto

-0.40

 escalada

-0.35

 Spitze

-0.35

hadapi

-0.35

 secund

-0.34

ActionButton

-0.34

ActionEvent

-0.34

 Wikiseite

-0.32

POSITIVE LOGITS

IntoConstraints

0.59

хьтан

0.56

 linkovi

0.56

 Италијани

0.52

 wireType

0.52

ftagPool

0.49

+#+#

0.48

 Enix

0.48

 ProtoMessage

0.46

Autoritní

0.45

Activations Density 0.033%

words and phrases related to decision-making and processes involving claims and reviews

will send email

No Comments

No Known Activations