INDEX

Explanations

references to malignant tumors and cancer-related terminology

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

malignant tumors or conditions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

DockStyle

-0.54

 Picchu

-0.54

demos

-0.50

 ErrIntOverflow

-0.48

 allAfrica

-0.47

torta

-0.46

ритори

-0.46

nologue

-0.45

ConverterFactory

-0.45

hnia

-0.45

POSITIVE LOGITS

 malignant

1.38

 malign

1.05

 benign

0.96

 malignancy

0.90

ignant

0.85

 malicious

0.61

 cancerous

0.56

 neop

0.53

ignants

0.50

 carcinogenic

0.50

Activations Density 0.017%

references to malignant tumors and cancer-related terminology

malignant tumors or conditions

No Comments

No Known Activations