INDEX

Explanations

references to original or authentic items and their characteristics

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

original or factory condition

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

Zusammenfassung

-0.53

enumii

-0.52

RegressionTest

-0.51

ChildScrollView

-0.50

RectangleBorder

-0.47

XmlAccessType

-0.47

Legături

-0.45

摘要

-0.45

 Вікіпе

-0.43

 ſche

-0.42

POSITIVE LOGITS

 original

0.66

 Original

0.59

Original

0.59

 originale

0.58

 ORIGINAL

0.58

 originais

0.57

 ویکی‌پدی

0.55

original

0.53

 originales

0.52

ORIGINAL

0.51

Activations Density 0.007%

references to original or authentic items and their characteristics

original or factory condition

No Comments

No Known Activations