INDEX

Explanations

references to design patterns and coding principles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ymce

-0.53

 étoit

-0.47

 privata

-0.47

Бахар

-0.47

posedge

-0.46

 doccia

-0.45

 vectorielle

-0.44

 pouvoit

-0.44

pagation

-0.43

 pubblicità

-0.43

POSITIVE LOGITS

patterns

0.51

 patterns

0.50

Glej

0.49

ValueStyle

0.46

prnewswire

0.45

lösung

0.42

 Italijanski

0.42

 solution

0.41

 Patterns

0.41

Patterns

0.40

Activations Density 0.008%

references to design patterns and coding principles

No Comments

No Known Activations