INDEX

Explanations

terms and concepts related to JSON and XML processing or conversion

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

data types and formats

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

 estekak

-0.65

 CreateTagHelper

-0.64

IntoConstraints

-0.63



-0.63

MemoryWarning

-0.63

المناصب

-0.61

mybatisplus

-0.61

 AssemblyTitle

-0.59

 հղումներ

-0.58

تقاوى

-0.57

POSITIVE LOGITS

 equivalent

0.32

 équi

0.28

equivalent

0.27

équi

0.26

 Pfund

0.26

 automáticamente

0.25

 פל

0.24

 Dafür

0.24

developer

0.24

 equivalente

0.24

Activations Density 0.093%

terms and concepts related to JSON and XML processing or conversion

data types and formats

No Comments

No Known Activations