INDEX

Explanations

geographic locations and infrastructure references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

RenderAtEndOf

-0.31

 neben

-0.30

圆

-0.30

 بوس

-0.30

ränkt

-0.29

 Рим

-0.29

tadora

-0.28

 hormone

-0.28

Dorf

-0.27

WriteLiteral

-0.27

POSITIVE LOGITS

 north

2.06

 south

2.05

 east

1.97

 west

1.94

 eastward

1.77

 eastern

1.73

 southern

1.71

 southward

1.70

 western

1.70

 northern

1.70

Activations Density 0.510%

geographic locations and infrastructure references

No Comments

No Known Activations