INDEX

Explanations

words associated with "relatedness" or connections among concepts

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

related [noun describing activity/outcome]

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

 HasFactory

-0.45

 juges

-0.44

 Inscrivez

-0.44

autonomie

-0.41

 posesión

-0.41

skjaer

-0.40

\{\\

-0.39

 autonomía

-0.39

 समीक्षक

-0.39

 voluntad

-0.38

POSITIVE LOGITS

related

0.96

 related

0.87

Related

0.83

 Related

0.75

 activities

0.70

 RELATED

0.65

associated

0.62

relevant

0.61

相关的

0.59

 activity

0.58

Activations Density 0.016%

words associated with "relatedness" or connections among concepts

related [noun describing activity/outcome]

No Comments

No Known Activations