INDEX

Explanations

technical terminology related to processes and data management in scientific or computational contexts

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

Preceding "parts", "DC", or "frequencies"

np_acts-logits-general · gemini-2.0-flash

acquired or newly completed items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

なもの

-0.41

neither

-0.41

something

-0.40

knew

-0.38

Took

-0.36

 spoke

-0.36

 لازم

-0.36

 majeurs

-0.35

Trag

-0.35

又是

-0.35

POSITIVE LOGITS

 newly

1.22

 repaired

1.10

 installed

1.09

 extracted

1.09

newly

1.08

 downloaded

1.05

 acquired

1.04

 freshly

1.03

 renovated

1.02

 purchased

1.02

Activations Density 1.460%

technical terminology related to processes and data management in scientific or computational contexts

Preceding "parts", "DC", or "frequencies"

acquired or newly completed items

No Comments

No Known Activations