INDEX

Explanations

symbols and markup elements used in coding or formatting

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

Celtic, NPR, Peugeot, Puccini

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

 незавершена

-0.52

 itself

-0.47

mbggenerated

-0.46

GEBURTSDATUM

-0.45

 яке

-0.45

his

-0.43

embar

-0.41

 gouvernements

-0.40

IActionResult

-0.40

POSITIVE LOGITS

 jotka

0.73

 themselves

0.65

 amelyek

0.64

 neler

0.63

themselves

0.61

 lisäksi

0.59

 generalizations

0.59

 αυτά

0.56

 които

0.55

 אלה

0.55

Activations Density 0.312%

symbols and markup elements used in coding or formatting

Celtic, NPR, Peugeot, Puccini

No Comments

No Known Activations