INDEX

Explanations

terms related to enabling and disabling functionalities or features within systems

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

Disable/ignore functionality

np_acts-logits-general · gemini-2.0-flash

disable/stop actions

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

-0.52

AutoField

-0.51

rungsseite

-0.50

IntoConstraints

-0.49

 Signalez

-0.47

 juſ

-0.46

 actu

-0.44

 eût

-0.43

+#+

-0.43

 AssemblyCompany

-0.42

POSITIVE LOGITS

 disable

0.60

disable

0.59

 disabling

0.55

 Disable

0.52

skip

0.51

override

0.51

 exemption

0.49

Disable

0.48

 shielding

0.47

EnableWeb

0.47

Activations Density 0.842%

terms related to enabling and disabling functionalities or features within systems

Disable/ignore functionality

disable/stop actions

No Comments

No Known Activations