INDEX

Explanations

technical terminology and notation within mathematical discussions

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

stating definitions or considering models

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

 also

-0.32

kt

-0.30

 online

-0.28

 وتسجيلات

-0.26

 flat

-0.26

 bill

-0.26

-0.25

 protection

-0.25

fastjson

-0.25

iprot

-0.24

POSITIVE LOGITS

ReusableCell

0.75

⤹

0.74

 otomatig

0.73

 kasarigan

0.73

 propOrder

0.72

msgTypes

0.71

 disambiguazione

0.70

 ModelExpression

0.70

 Administrativna

0.68

Personendaten

0.64

Activations Density 0.188%

technical terminology and notation within mathematical discussions

stating definitions or considering models

No Comments

No Known Activations