INDEX

Explanations

instances of user interaction or expressions that indicate a request for help or guidance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 vostro

-1.03

 לכם

-0.97

 لكم

-0.96

 vostra

-0.91

 yourselves

-0.89

您

-0.87

 vostri

-0.86

你們

-0.86

你们

-0.84

 yall

-0.83

POSITIVE LOGITS

 thou

1.11

thy

1.03

 ты

0.89

 твои

0.85

Thy

0.80

 Thou

0.80

 твое

0.79

 тебе

0.78

Thy

0.78

Ты

0.78

Activations Density 0.198%

instances of user interaction or expressions that indicate a request for help or guidance

No Comments

No Known Activations

instances of user interaction or expressions that indicate a request for help or guidance

No Comments

No Known Activations