INDEX

Explanations

punctuation marks and conjunctions within lists or enumerations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AutoScale

-0.49

antino

-0.47

請繼續往下閱讀

-0.44

zinger

-0.43

zaamheid

-0.43

ſelves

-0.42

MLLoader

-0.41

 utafitiHapana

-0.41

 bezeichneter

-0.41

literal

-0.41

POSITIVE LOGITS

="//

0.38

 pétrole

0.38

 ModelExpression

0.38

 AttributeSet

0.37

цездатний

0.36

 poveznice

0.35

 traffico

0.35

ChildScrollView

0.35

 requestCode

0.34

 regra

0.34

Activations Density 0.038%

punctuation marks and conjunctions within lists or enumerations

No Comments

No Known Activations