INDEX

Explanations

connects multiple parts of a legal argument.

oai_token-act-pair · gemini-2.0-flash

for/that

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_4/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 urges

-0.48

 urge

-0.48

to

-0.47

 urged

-0.45

你先

-0.43

cast

-0.42

beln

-0.40

 people

-0.40

 motivated

-0.39

тия

-0.39

POSITIVE LOGITS

CppMethod

0.95

 ویکی‌پدی

0.83

 насељу

0.82

 Приступљено

0.81

]")]

0.81

^(@)

0.81

 initComponents

0.79

']")

0.75

 дописавши

0.74

 Roskov

0.74

Activations Density 2.782%

connects multiple parts of a legal argument.

for/that

No Comments

No Known Activations