INDEX

Explanations

phrases related to numerical values and patient counts in clinical studies

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

patients closed

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 MediatR

-0.58

 OFDb

-0.51

RTLR

-0.50

 Roskov

-0.50

 endblock

-0.48

 viewDidLoad

-0.48

IntoConstraints

-0.48

sizeCache

-0.48

 cherchés

-0.47

unknownFields

-0.45

POSITIVE LOGITS

 antaranya

0.48

 ramah

0.39

󠁣

0.38

 glisser

0.38

 diantaranya

0.37

hilangan

0.36

 recev

0.36

 korun

0.36

 Chwiliwch

0.35

 derechos

0.35

Activations Density 0.107%

phrases related to numerical values and patient counts in clinical studies

patients closed

No Comments

No Known Activations