INDEX

Explanations

references to previously established works and comparisons to other products or publications

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

many earlier attempts

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

parsedMessage

-0.61

awtextra

-0.61

RegressionTest

-0.59

⤹

-0.59

yntaxException

-0.57

IBOutlet

-0.57

MessageTagHelper

-0.56

IntoConstraints

-0.55

脚注の使い方

-0.52

AddHtmlAttribute

-0.52

POSITIVE LOGITS

 merely

0.41

 ordinary

0.39

 Banks

0.37

 many

0.36

 popular

0.35

 comum

0.34

But

0.34

few

0.33

 simply

0.32

 mere

0.32

Activations Density 0.057%

references to previously established works and comparisons to other products or publications

many earlier attempts

No Comments

No Known Activations