INDEX

Explanations

specific references to studies and evaluations, particularly in relation to classification and regression methods in research

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

saraba

-0.59

XmlAccessorType

-0.58

GIVEREF

-0.56

tonsoft

-0.54

 oprot

-0.53

 onOptions

-0.53

onAttach

-0.53



-0.52

 AssemblyTitle

-0.52

EDEFAULT

-0.51

POSITIVE LOGITS

 disposiciones

0.36

transQ

0.33

 bebes

0.32

蚀

0.31

weilen

0.31

 ressources

0.30

 weshalb

0.29

mitteln

0.29

ногда

0.29

 occasionally

0.28

Activations Density 0.148%

specific references to studies and evaluations, particularly in relation to classification and regression methods in research

No Comments

No Known Activations