INDEX

Explanations

phrases that express uncertainty or conditionality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

клопе

-0.91

ergies

-0.86

ECA

-0.86

SBATCH

-0.85

 تضيفلها

-0.84

ollectionView

-0.84

quelize

-0.84

 kasarigan

-0.78

 Breen

-0.78

 Cassel

-0.77

POSITIVE LOGITS

 might

1.95

might

1.85

Might

1.80

 Might

1.79

 MIGHT

1.68

 pourraient

0.94

 could

0.93

 podría

0.90

 pourrait

0.89

ght

0.89

Activations Density 0.064%

phrases that express uncertainty or conditionality

No Comments

No Known Activations