INDEX

Explanations

closure or termination markers in programming syntax

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

code and technical terms

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

ſicht

-0.68

RTLR

-0.66

 queſta

-0.66

 ſei

-0.66

ſehen

-0.66

 ſeine

-0.65

 dieſer

-0.65

niſſe

-0.63

 laſſen

-0.63

 beſch

-0.62

POSITIVE LOGITS

StoryboardSegue

0.33

 Silber

0.29

ArgumentParser

0.29

 simak

0.28

 Relaciones

0.28

runApp

0.28

 ustedes

0.27

 naranja

0.27

modelBuilder

0.26

 informado

0.26

Activations Density 0.005%

closure or termination markers in programming syntax

code and technical terms

No Comments

No Known Activations