INDEX

Explanations

relationships involving groups or hierarchies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

一身

-0.34

套

-0.33

 someone

-0.32

的一个

-0.32

一口

-0.32

 reality

-0.32

好

-0.31

 stimulation

-0.30

 somebody

-0.30

 enemy

-0.30

POSITIVE LOGITS

Datuak

0.87

Tikang

0.79

AddTagHelper

0.77

 BoxDecoration

0.72

Personendaten

0.71

::{

0.71

styleUrls

0.71

oneofs

0.70

LookAnd

0.69

 propOrder

0.69

Activations Density 0.904%

relationships involving groups or hierarchies

No Comments

No Known Activations