INDEX

Explanations

uses of the words "will", "would", "going to", "able to", and forms of "be" along with personal pronouns, suggesting it detects predictions or intentions.

oai_token-act-pair · gemini-2.0-flash

Action words

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_23/width_16k/average_l0_25

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

 giudi

-0.65

<bos>

-0.62

 itſelf

-0.60

 pulito

-0.57

 titolata

-0.57

 calda

-0.57

 manutenzione

-0.57

 følgende

-0.57

 kunder

-0.56

 NDEBUG

-0.56

POSITIVE LOGITS

be

0.78

 guide

0.68

 rule

0.68

 document

0.64

KURZBESCHREIBUNG

0.61

]();

0.61

 pass

0.60

 capture

0.58

 admit

0.58

 write

0.58

Activations Density 7.385%

uses of the words "will", "would", "going to", "able to", and forms of "be" along with personal pronouns, suggesting it detects predictions or intentions.

Action words

No Comments

No Known Activations

uses of the words "will", "would", "going to", "able to", and forms of "be" along with personal pronouns, suggesting it detects predictions or intentions.

Action words

No Comments

No Known Activations