INDEX

Explanations

attends to symbols and numbers from code-related tokens

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.09

1:0.10

2:0.12

3:0.10

4:0.05

5:0.02

6:0.23

7:0.24

Negative Logits

fjspx

-0.41

--]

-0.37

存于互联网档案馆

-0.32

 Ruman

-0.30

 MonoBehaviour

-0.30

CELLANEOUS

-0.29

phonie

-0.29

});*/

-0.28

دانشنامهٔ

-0.28

 ==========

-0.28

POSITIVE LOGITS

ItemBackground

0.35

contentLoaded

0.30

 animés

0.30

zaki

0.29

rungsseite

0.29

 بيها

0.28

 Banken

0.28

rated

0.28

 Round

0.28

picasso

0.28

Activations Density 0.003%

attends to symbols and numbers from code-related tokens

No Comments

No Known Activations