INDEX

Explanations

references to intensive care and related medical terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 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

 avoient

-0.41

래스

-0.39

 magnética

-0.36

 générations

-0.35

 almuerzo

-0.35

ấm

-0.35

 inégal

-0.35

 pubb

-0.34

 étoient

-0.34

 大全

-0.34

POSITIVE LOGITS

ICU

0.93

ICU

0.85

 Intensive

0.72

 preterm

0.70

 neonatal

0.70

 premature

0.69

 ventilator

0.67

 intensive

0.64

 Neon

0.62

 intensiv

0.60

Activations Density 0.657%

references to intensive care and related medical terminology

No Comments

No Known Activations