INDEX

Explanations

numerical values and abbreviations related to programming or technical specifications

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

numerical and code identifiers

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

ându

-0.36

 regulador

-0.35

 MATTERS

-0.35

 Wörter

-0.34

 Matters

-0.34

SEGUIR

-0.33

 amorti

-0.32

 Hinweise

-0.31

 ENDS

-0.31

 matters

-0.30

POSITIVE LOGITS

 utafitiHapana

0.78

 resourceCulture

0.76

 препратки

0.72

 kaarangay

0.71

 gynhyrchwyd

0.71

 Wikimedijinoj

0.70

 autorytatywna

0.69

цездатний

0.66

TintMode

0.65

oa̍t

0.65

Activations Density 0.404%

numerical values and abbreviations related to programming or technical specifications

numerical and code identifiers

No Comments

No Known Activations