INDEX

Explanations

instances where a document structure is initiated, particularly in programming or code contexts

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

after open parentheses

np_acts-logits-general · gemini-2.0-flash

place names and people

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-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

.”

-0.53

</em>

-0.52

 when

-0.52

 what

-0.52

the

-0.52

“

-0.51

Do

-0.50

-0.49

</strong>

-0.49

POSITIVE LOGITS

 bezeichneter

0.93

 المعيارى

0.92

OGND

0.90

BeginContext

0.85

ScopeManager

0.83

 pseudonym

0.82

ⓘ

0.81

SBATCH

0.79

verwijspagina

0.79

Personensuche

0.79

Activations Density 0.141%

instances where a document structure is initiated, particularly in programming or code contexts

after open parentheses

place names and people

No Comments

No Known Activations