INDEX

Explanations

terms related to preferences and predominance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 rightly

-0.47

 cảnh

-0.47

 duly

-0.46

izarre

-0.44

 suivante

-0.43

 ويكي

-0.43

 TOUCH

-0.43

Ouch

-0.42

 tardi

-0.41

кви

-0.41

POSITIVE LOGITS

 Prefer

1.52

prefer

1.51

 prefer

1.41

Prefer

1.41

 preferred

1.40

 préfé

1.36

 Preference

1.30

 prefers

1.29

 Preferred

1.28

 preference

1.27

Activations Density 0.424%

terms related to preferences and predominance

No Comments

No Known Activations

terms related to preferences and predominance

No Comments

No Known Activations