INDEX

Explanations

references to explosive devices or weapons, particularly bombs

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

Cooling

-0.55

Архівовано

-0.53

kton

-0.52

EndInit

-0.51

BufferException

-0.50

 Cooling

-0.49

haustible

-0.49

writeFieldEnd

-0.49

SequentialGroup

-0.49

 messes

-0.48

POSITIVE LOGITS

cub

0.69

ub

0.66

latitude

0.65

Lat

0.64

lat

0.64

lat

0.63

Cub

0.62

Cub

0.62

Lat

0.60

 lateral

0.60

Activations Density 0.433%

references to explosive devices or weapons, particularly bombs

No Comments

No Known Activations