INDEX

Explanations

patterns related to numerical and coded values

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

Hexadecimal numbers

np_acts-logits-general · gemini-2.0-flash

key: value pairs

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

vna

-0.47

Mar

-0.47

ag

-0.45

Ch

-0.44

Portály

-0.43

 regional

-0.43

utta

-0.43

وائ

-0.42

POSITIVE LOGITS

Personensuche

0.84

دانشنامهٔ

0.79

DoubleQuotes

0.78

 itſelf

0.78

 pleaſure

0.74

extAlignment

0.73

 esternos

0.73

 виправивши

0.73

脚注の使い方

0.72

 themſelves

0.72

Activations Density 0.600%

patterns related to numerical and coded values

Hexadecimal numbers

key: value pairs

No Comments

No Known Activations