INDEX

Explanations

references to websites and online resources for information

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

find more details or pages

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

SourceChecksum

-0.40

 Италијани

-0.39

جغرافيا

-0.39

ⓧ

-0.36

limsy

-0.36

 Kommission

-0.35

 Relaciones

-0.35

ByVersion

-0.35

 múl

-0.35

 Egipto

-0.35

POSITIVE LOGITS

########.

0.56

 HomeController

0.48

 jsPsych

0.48

 المعيارى

0.46

glades

0.46

 memoized

0.45

 referrerpolicy

0.45

 doInBackground

0.44

EndInit

0.44

 XCTest

0.43

Activations Density 0.016%

references to websites and online resources for information

find more details or pages

No Comments

No Known Activations