INDEX

Explanations

code structuring elements such as comments and function marks

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

user commands and code structure

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

pecabe

-0.80

ロウィン

-0.78

uxxxx

-0.76

 дописавши

-0.74

Tikang

-0.73

出版年

-0.72

+#+#

-0.71

writeField

-0.70

ésultats

-0.69

 ویکی‌پدی

-0.69

POSITIVE LOGITS

0.41

0.38

0.37

0.35

↵↵

0.35

 Next

0.35

Related

0.34

↵

0.33

Activations Density 0.001%

code structuring elements such as comments and function marks

user commands and code structure

No Comments

No Known Activations