INDEX

Explanations

the use of arrows or indicators in formatted text or code snippets

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

arrows or separators

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

ssymb

-0.45

 añadido

-0.40

-0.39

CppMethod

-0.38

 encontraba

-0.36

/*

-0.35

 británico

-0.35

Servico

-0.35

 Naciones

-0.35

 inoxidable

-0.35

POSITIVE LOGITS

rightarrow

0.91

sendRedirect

0.87

 --->

0.86

//-->

0.84

Rightarrow

0.84

-->

0.83

-->

0.81

→

0.79

 –>

0.78

→

0.77

Activations Density 0.076%

the use of arrows or indicators in formatted text or code snippets

arrows or separators

No Comments

No Known Activations