INDEX

Explanations

patterns in structured data or code, particularly focusing on numerical sequences or structured formats

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

numeric codes or identifiers

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

code structure numbers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

脚注の使い方

-0.56

RTLI

-0.52

 وتسجيلات

-0.49

 تعدى

-0.49

 الدولى

-0.47

gatsby

-0.44

 almost

-0.44

См

-0.43

 norma

-0.43

die

-0.43

POSITIVE LOGITS

 дописавши

0.96

 '\\;'

0.85

 Paglinawan

0.83

\{\\

0.83

MigrationBuilder

0.81

 nakalista

0.80

RenderAtEndOf

0.76

()]);

0.76

 ویکی‌پدی

0.75

 <<<<<<<<<<<<<<

0.75

Activations Density 0.211%

patterns in structured data or code, particularly focusing on numerical sequences or structured formats

numeric codes or identifiers

code structure numbers

No Comments

No Known Activations

patterns in structured data or code, particularly focusing on numerical sequences or structured formats

numeric codes or identifiers

code structure numbers

No Comments

No Known Activations