INDEX

Explanations

mentions of exposure to risks or potential harm

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

(!__

-0.66

bitField

-0.66

httphttps

-0.65

 lubrication

-0.63

('/:

-0.62

 Melch

-0.62

 Granger

-0.62

 thickening

-0.62

serviceWorker

-0.60

Adel

-0.60

POSITIVE LOGITS

 exposure

2.79

 Exposure

2.59

 exposed

2.54

exposure

2.51

 expose

2.49

 exposing

2.43

 Exposed

2.40

Exposure

2.40

Exposed

2.38

exposed

2.37

Activations Density 0.182%

mentions of exposure to risks or potential harm

No Comments

No Known Activations