INDEX

Explanations

terms related to preparation and planning ahead

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

namon

-0.63

)))));

-0.61

enetrable

-0.60

Portály

-0.57

gdx

-0.53

 testGet

-0.51

 Mand

-0.50

—

-0.50

ailangan

-0.50

etc

-0.50

POSITIVE LOGITS

 beforehand

1.07

 заранее

0.99

 antemano

0.98

 vooraf

0.96

pre

0.94

 preload

0.93

提前

0.89

 ahead

0.87

 previo

0.85

 pré

0.83

Activations Density 0.294%

terms related to preparation and planning ahead

No Comments

No Known Activations