INDEX

Explanations

instances of first person pronouns and the copula "to be" used with modals.

oai_token-act-pair · gemini-2.0-flash

forms of the verb "to be"

np_token-act-pair-logits · gpt-4o-mini

"to be"

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

ніципалі

-0.72

 EconPapers

-0.71

IUrlHelper

-0.69

MessageOf

-0.67

viewDidLoad

-0.63

setVerticalGroup

-0.63

Rüyada

-0.62

 Monfieur

-0.60

 CURIAM

-0.60

 CanadaChoose

-0.59

POSITIVE LOGITS

is

0.82

be

0.81

are

0.77

Be

0.64

 быть

0.63

Are

0.63

 jesteśmy

0.62

 być

0.61

NotBe

0.60

 jestem

0.59

Activations Density 2.108%

instances of first person pronouns and the copula "to be" used with modals.

forms of the verb "to be"

"to be"

No Comments

No Known Activations