INDEX

Explanations

references to bans or prohibitions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 referenties

-0.83

thâu

-0.83

 Administrativna

-0.81

contentLoaded

-0.75

énario

-0.70

مصادر

-0.69

 transférez

-0.68

 Wikimedijinoj

-0.68

elance

-0.67

 accomp

-0.67

POSITIVE LOGITS

 banned

2.00

ban

1.88

 bans

1.85

 prohibited

1.76

 prohibit

1.71

 prohibition

1.69

 banning

1.64

 forbidden

1.55

banned

1.54

 forbid

1.47

Activations Density 0.104%

references to bans or prohibitions

No Comments

No Known Activations