INDEX

Explanations

references to patents or patent applications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

CloseOperation

-0.44

 princí

-0.40

 samband

-0.40

 budeme

-0.40

 eléctrico

-0.40

 normalidad

-0.40

 sagrado

-0.40

gameserver

-0.39

 académica

-0.39

 culturel

-0.38

POSITIVE LOGITS

 kasarigan

0.74

➟

0.51

Hochspringen

0.49

$.}

0.48

}));

0.47

 Cassini

0.47

')]

0.47

$.

0.46

❋

0.46

__':

0.46

Activations Density 0.004%

references to patents or patent applications

No Comments

No Known Activations