INDEX

Explanations

punctuation marks and specific numeric values

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

timestamps and punctuation

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

 препратки

-0.66

ſelf

-0.61

 ſei

-0.58

 auffi

-0.57

 Verſ

-0.56

 Majefty

-0.56

iffance

-0.55

 ſind

-0.55

ésultats

-0.54

 geſch

-0.54

POSITIVE LOGITS

 pinulongan

0.43

Personendaten

0.37

Literatur

0.35

 StatefulWidget

0.35

MOCK

0.35

onAttach

0.33

Externé

0.32

Personensuche

0.31

Ha

0.31

ponses

0.31

Activations Density 0.002%

punctuation marks and specific numeric values

timestamps and punctuation

No Comments

No Known Activations