INDEX

Explanations

terms related to comfort and support

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

下载附件

-0.41

 vektör

-0.41

berdayakan

-0.39

 bendera

-0.38

์ตูน

-0.38

 tanggung

-0.38

羯

-0.37

 państw

-0.37

 Álvarez

-0.36

Pó

-0.36

POSITIVE LOGITS

 comfort

0.81

Comfort

0.80

comfort

0.80

comfortable

0.76

 COMFORT

0.75

 comfortable

0.75

ChildScrollView

0.75

 Comfort

0.73

Comfortable

0.73

 comfy

0.72

Activations Density 0.005%

terms related to comfort and support

No Comments

No Known Activations