INDEX

Explanations

phrases related to confidentiality and information security

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

bringung

-0.51

Pageable

-0.48

 חיצוניים

-0.47

 humaine

-0.46

एँ

-0.43

 bakar

-0.42

 surat

-0.42

 contribué

-0.41

 Angebot

-0.41

 Enroll

-0.41

POSITIVE LOGITS

 secrecy

1.09

 Confidential

1.00

 confidentiality

0.99

 Confidentiality

0.92

 confidential

0.91

 secrets

0.87

Confidential

0.85

identiality

0.83

 Secrets

0.81

Secrets

0.79

Activations Density 0.347%

phrases related to confidentiality and information security

No Comments

No Known Activations