INDEX

Explanations

code-related symbols and syntax

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 queſta

-0.76

 Photocase

-0.75

 ſind

-0.72

 ddelweddau

-0.70

 كومونز

-0.69

 يتيمه

-0.68

onedDateTime

-0.67

ロウィン

-0.66

ỡng

-0.66

 Мексичка

-0.65

POSITIVE LOGITS

Po

0.29

GR

0.29

 Party

0.29

 Dank

0.29

Props

0.28

IS

0.28

 your

0.28

getClass

0.28

 tired

0.28

Spe

0.28

Activations Density 0.042%

code-related symbols and syntax

No Comments

No Known Activations