INDEX

Explanations

terms related to legal liability and conditions

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

refundable

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



-0.56

stateProvider

-0.50

]=>

-0.49

 utafitiHapana

-0.47

DockStyle

-0.41

UrlResolution

-0.40

KommentareTeilen

-0.37

ExtendWith

-0.37

AnimationsModule

-0.36

çon

-0.36

POSITIVE LOGITS

 refundable

3.30

refundable

2.56

 Refund

1.32

Refund

1.30

 REFUND

1.28

 refund

1.23

 refunded

1.23

 refunds

1.13

refund

1.11

REFUND

1.07

Activations Density 0.000%

terms related to legal liability and conditions

refundable

No Comments

No Known Activations