INDEX

Explanations

references to links and instructions for accessing content

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

click through

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Öffentlichkeit

-0.47

 Notwendigkeit

-0.37

 Konsequ

-0.36

 Voraussetzungen

-0.35

 evidence

-0.35

 Existenz

-0.34

 Gleichzeitig

-0.34

 hoor

-0.33

 consequences

-0.33

άνι

-0.33

POSITIVE LOGITS

IUrlHelper

0.77

sizeCache

0.77

webElementXpaths

0.73



0.64

DockStyle

0.63

 مشين

0.63

ThroughAttribute

0.63

 мәкал

0.63

IntoConstraints

0.62

NameInMap

0.60

Activations Density 0.039%

references to links and instructions for accessing content

click through

No Comments

No Known Activations