INDEX

Explanations

attends to display-related tokens from gear-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

2:0.10

3:0.11

4:0.10

5:0.06

6:0.20

7:0.20

Negative Logits

zeczy

-0.26

 уда

-0.26

 beschik

-0.25

 POSS

-0.25

 idea

-0.25

"[

-0.25

Else

-0.24

勤

-0.24

fair

-0.24

POSITIVE LOGITS

 CreateTagHelper

0.41

pozdrawiam

0.35

 actionMode

0.34

醐

0.33

Życiorys

0.33

 OMITBAD

0.32

 insuffisamment

0.31

 initComponents

0.31

izziness

0.30

 '\\;'

0.30

Activations Density 0.066%

attends to display-related tokens from gear-related tokens

No Comments

No Known Activations