INDEX

Explanations

works

np_token-act-pair-logits · gemini-2.0-flash

words and phrases indicating tasks being successfully completed

oai_token-act-pair · gemini-2.0-flash

Functionality

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_16/width_16k/average_l0_10

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.16.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 EdgeInsets

-0.63

ViewFeatures

-0.59

posedge

-0.58

CppMethod

-0.58

Rüyada

-0.57

bootstrapcdn

-0.55

 nahilalakip

-0.55

 <<<<<<<<<<<<<<

-0.55

 transfieras

-0.54

RSpec

-0.52

POSITIVE LOGITS

таратура

0.51

 CRUZ

0.49

arate

0.48

świet

0.47

 fantaisie

0.46

 Parcelable

0.46

ased

0.46

jectable

0.45

 spare

0.45

 rezerv

0.44

Activations Density 0.941%

works

words and phrases indicating tasks being successfully completed

Functionality

No Comments

No Known Activations