INDEX

Explanations

words and phrases indicating something new, especially in education or politics

oai_token-act-pair · gemini-2.0-flash

Opinions and consequences

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 épaules

-0.84

 travailleurs

-0.73

 esprits

-0.73

 touristes

-0.71

 Juifs

-0.70

 prisonniers

-0.69

 lèvres

-0.69

 dieux

-0.67

 consommateurs

-0.66

 âmes

-0.65

POSITIVE LOGITS

SequentialGroup

0.83

 ModelExpression

0.78



0.77

期刊论文

0.74

MENAFN

0.69

Hauptartikel

0.68

SBATCH

0.67

ConstraintMaker

0.67

OGND

0.66

 تانيه

0.66

Activations Density 17.273%

words and phrases indicating something new, especially in education or politics

Opinions and consequences

No Comments

No Known Activations