INDEX

Explanations

specific patterns or formats related to citations or references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

 незавершена

-0.84

 الحره

-0.76

Datuak

-0.64

ècie

-0.64

tagHelperRunner

-0.63

Kjelder

-0.62

Portale

-0.60

Hauptartikel

-0.57

تقاوى

-0.56

pensione

-0.55

POSITIVE LOGITS

adaptiveStyles

0.53

styleType

0.50

 QtCore

0.50

EndInit

0.49

ilang

0.45

Dtor

0.45

ogyn

0.44

XmlEnum

0.43

◀

0.42

ikyuu

0.42

Activations Density 0.043%

specific patterns or formats related to citations or references

No Comments

No Known Activations