INDEX

Explanations

attends to asterisks denoting sequences of indexing changes from the preceding tokens, suggesting it focuses on modifications in code or configuration files

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

0:0.08

1:0.08

2:0.49

3:0.11

4:0.05

5:0.02

6:0.05

7:0.08

Negative Logits

EndContext

-0.44

celes

-0.29

 Hano

-0.26

 señores

-0.26

 épis

-0.25

 verità

-0.25

 Obed

-0.25

utuhkan

-0.25

ölkerung

-0.24

 <=",

-0.24

POSITIVE LOGITS

 فريبيس

0.28

 INTERESAR

0.27

autoconfigure

0.25

UrlResolution

0.25

SnackBar

0.25

 &___

0.24

omonas

0.24

 portato

0.24

ztyn

0.23

 both

0.23

Activations Density 0.280%

attends to asterisks denoting sequences of indexing changes from the preceding tokens, suggesting it focuses on modifications in code or configuration files

No Comments

No Known Activations