INDEX

Explanations

highly numerical and specific patent-related information

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

U.S. Pat. No

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

ulement

-0.34

cuál

-0.33

 watching

-0.33

etton

-0.32

 ranchers

-0.32

morgan

-0.31

 Pakistani

-0.31

sibly

-0.31

asiness

-0.31

POSITIVE LOGITS

tagHelperRunner

0.82

RegressionTest

0.69

AccessorTable

0.63

findpost

0.63

uxxxx

0.62

Geplaatst

0.61

InjectAttribute

0.60

хьтан

0.60

adaptiveStyles

0.60

LabelTagHelper

0.59

Activations Density 0.017%

highly numerical and specific patent-related information

U.S. Pat. No

No Comments

No Known Activations