INDEX

Explanations

necessity and requirement-related terms, indicating what needs to be done or is essential

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

be

-0.59

InputBorder

-0.45

 puissent

-0.45

 aient

-0.42

 parezca

-0.41

 puisse

-0.39

 كومونز

-0.39

XCTAssert

-0.36

 chrétiens

-0.36

 podamos

-0.35

POSITIVE LOGITS

 will

0.71

########.

0.57

存于互联网档案馆

0.56

 Италијани

0.54

 станет

0.54

will

0.52

期刊论文

0.52

erà

0.50

IntoConstraints

0.50

arán

0.49

Activations Density 0.390%

necessity and requirement-related terms, indicating what needs to be done or is essential

No Comments

No Known Activations