INDEX

Explanations

references to time periods, specifically years or decades

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

decade

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

 mohl

-0.38

CELLANEOUS

-0.35

 wenigen

-0.35

ěli

-0.34

 wenige

-0.34

posix

-0.33

 stanu

-0.32

omoto

-0.32

 Spaces

-0.32

zí

-0.32

POSITIVE LOGITS

 years

0.81

decade

0.77

years

0.76

YEARS

0.68

 década

0.68

Years

0.67

decade

0.65

 Years

0.64

 YEARS

0.64

 años

0.63

Activations Density 0.017%

references to time periods, specifically years or decades

decade

No Comments

No Known Activations

references to time periods, specifically years or decades

decade

No Comments

No Known Activations