INDEX

Explanations

references to measurements or assessments involving probes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 civilización

-0.57

 regalías

-0.48

 meditación

-0.47

 religión

-0.47

 naciones

-0.47

 conexiones

-0.47

 akrab

-0.46

 atracción

-0.46

 fronteras

-0.45

 ejecut

-0.45

POSITIVE LOGITS

 protoimpl

0.77

 probes

0.72

 probe

0.70

mal

0.69

boot

0.68

 probing

0.66

 Probe

0.66

 Coffee

0.63

Coffee

0.62

ImageContext

0.62

Activations Density 0.208%

references to measurements or assessments involving probes

No Comments

No Known Activations