INDEX

Explanations

mathematical notations and operations related to matrix algebra and projections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

Reasonable

-0.56

 beitragen

-0.55

 المعيارى

-0.52

abstractmethod

-0.52

 digress

-0.52

trailing

-0.50

enumii

-0.50

celona

-0.50

ثة

-0.49

ivant

-0.48

POSITIVE LOGITS

thâu

0.70

localctx

0.63

 cherchés

0.63

mybatisplus

0.63

 يتيمه

0.63

stdc

0.61

FXMLLoader

0.60

BeginContext

0.58

TypedDataSet

0.57

ExecuteReader

0.56

Activations Density 0.690%

mathematical notations and operations related to matrix algebra and projections

No Comments

No Known Activations