INDEX

Explanations

phrases indicating surprise or unexpected events

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

 BoxFit

-0.60

 estekak

-0.56

 &___

-0.53

########.

-0.53

Portale

-0.50

AnchorTagHelper

-0.49

ftagPool

-0.47

 TestBed

-0.45

 AssemblyTitle

-0.44

bosity

-0.44

POSITIVE LOGITS

 surprise

3.53

 surprises

3.03

 surprised

2.98

surprise

2.92

Surprise

2.88

 Surprise

2.84

 surprising

2.75

 Überras

2.58

 sorpresa

2.58

surprised

2.44

Activations Density 0.454%

phrases indicating surprise or unexpected events

No Comments

No Known Activations