INDEX

Explanations

references to meetings

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

about meetings

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

 trasparente

-0.65

 škoda

-0.58

 dieux

-0.57

 transparan

-0.54

 ecológico

-0.52

 digitais

-0.51

 ecológica

-0.50

 graça

-0.50

 transparencia

-0.49

 esportivo

-0.49

POSITIVE LOGITS

 Meeting

1.15

 meetings

1.10

 meeting

1.10

meeting

1.09

Meeting

1.09

 Meetings

1.08

 MEETING

0.97

meetings

0.93

Meetings

0.92

ETING

0.85

Activations Density 0.009%

references to meetings

about meetings

No Comments

No Known Activations