INDEX

Explanations

attends to programming language syntax tokens from language keywords or modifiers

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

0:0.04

1:0.08

2:0.05

3:0.04

4:0.22

5:0.46

6:0.03

7:0.04

Negative Logits

$")

-0.34

 tartalomajánló

-0.27

 للمعارف

-0.26

 Weit

-0.25

řevě

-0.24

+')

-0.24

"…

-0.24

ujudkan

-0.24

 erős

-0.24

 <=",

-0.23

POSITIVE LOGITS

parsedMessage

0.30

onde

0.28

SafeMath

0.27

writeFieldEnd

0.26

Zij

0.26

elemField

0.26

encio

0.25

zhong

0.25

LLocation

0.25

 Sandman

0.24

Activations Density 0.233%

attends to programming language syntax tokens from language keywords or modifiers

No Comments

No Known Activations