INDEX

Explanations

symbols and syntax related to formatting and code structures

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

Follows "</end_of_turn>"

np_acts-logits-general · gemini-2.0-flash

Corona, Google, Parse, Android

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Taktlose

-0.87

 snippetHide

-0.82

Datuak

-0.82

 autorytatywna

-0.71

 CanadaChoose

-0.68

Hentet

-0.67

GEBURTSDATUM

-0.66

 Normdatei

-0.66

 invokingState

-0.65

Vidite

-0.64

POSITIVE LOGITS

 Linder

0.36

je

0.35

Functional

0.32

 Part

0.32

anin

0.32

 Reward

0.32

lab

0.31

NO

0.31

 Risk

0.31

Activations Density 0.055%

symbols and syntax related to formatting and code structures

Follows "</end_of_turn>"

Corona, Google, Parse, Android

No Comments

No Known Activations

symbols and syntax related to formatting and code structures

Follows "</end_of_turn>"

Corona, Google, Parse, Android

No Comments

No Known Activations