INDEX

Explanations

programming-related keywords and structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 ſtate

-0.62

 Diſ

-0.59

 itſelf

-0.57

 Shakspeare

-0.57

 faſt

-0.56

 Vorm

-0.56

دانشنامهٔ

-0.54

ſelves

-0.52

 utafitiHapana

-0.50

 ftate

-0.50

POSITIVE LOGITS

 propOrder

0.85

 esternos

0.60

Personendaten

0.60

échal

0.57

 invokingState

0.55

Viitteet

0.55

الإنجليزية

0.54

 nhật

0.53

berton

0.53

 &___

0.52

Activations Density 0.334%

programming-related keywords and structures

No Comments

No Known Activations