INDEX

Explanations

keywords and concepts related to documentation, legal proceedings, and formal agreements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 především

-0.35

 autant

-0.33

舺

-0.32

 Especially

-0.31

Addo

-0.30

 especialmente

-0.30

ritter

-0.29

 Particularly

-0.29

 unterwegs

-0.29

 buurt

-0.29

POSITIVE LOGITS

 describing

1.70

 indicating

1.58

 explaining

1.56

 depicting

1.54

 detailing

1.53

 outlining

1.48

 showing

1.47

 announcing

1.41

 denoting

1.40

 stating

1.38

Activations Density 2.048%

keywords and concepts related to documentation, legal proceedings, and formal agreements

No Comments

No Known Activations