INDEX

Explanations

references to patents and their applications

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

patent application and journal citations

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

↵↵

-0.37

↵

-0.30

-0.29

 sedangkan

-0.29

 Dennoch

-0.28

<eos>

-0.28

 while

-0.28

However

-0.28

 แดง

-0.27

POSITIVE LOGITS

ngdoc

0.91

enderror

0.83

CppMethod

0.82

脚注の使い方

0.82

GOTREF

0.80

DockStyle

0.75

 propOrder

0.75

 تانيه

0.75

RTGC

0.75

RTLR

0.74

Activations Density 0.124%

references to patents and their applications

patent application and journal citations

No Comments

No Known Activations