INDEX

Explanations

attends to specific programming or technical keywords from configuration-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

0:0.10

1:0.17

2:0.12

3:0.11

4:0.15

5:0.07

6:0.11

7:0.14

Negative Logits

 незавершена

-0.61

drawSprites

-0.45

reactstrap

-0.40

 defaultstate

-0.40

 Signalez

-0.40

RenderAtEndOf

-0.40

 itſelf

-0.40

};*/

-0.40

SourceChecksum

-0.39

}$

-0.39

POSITIVE LOGITS

価値

0.23

 interesting

0.23

getJSONArray

0.22

<sup>

0.22

 career

0.22

<blockquote>

0.22

amin

0.22

And

0.21

dien

0.21

Activations Density 0.002%

attends to specific programming or technical keywords from configuration-related tokens

No Comments

No Known Activations