INDEX

Explanations

references to elements in a document or code, especially in the context of programming or structural analysis

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

"element" or "elements"

np_acts-logits-general · gemini-2.0-flash

elements of something

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 GenerationType

-0.52

 giù

-0.48

 Turquía

-0.48

 محفوظة

-0.48

 fiscale

-0.48

 Suiza

-0.47

 onStart

-0.45

 незавершена

-0.45

 Moscú

-0.44

 Kriegs

-0.43

POSITIVE LOGITS

 Element

1.36

Element

1.31

 element

1.28

element

1.20

 Elements

1.13

ElementException

1.13

 elements

1.11

ELEMENT

1.10

elements

1.07

 ELEMENT

1.05

Activations Density 0.095%

references to elements in a document or code, especially in the context of programming or structural analysis

"element" or "elements"

elements of something

No Comments

No Known Activations

references to elements in a document or code, especially in the context of programming or structural analysis

"element" or "elements"

elements of something

No Comments

No Known Activations