INDEX

Explanations

attends to tokens related to the expressions and commands from programming or technical contexts

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

0:0.14

1:0.15

2:0.09

3:0.08

4:0.08

5:0.16

6:0.14

7:0.13

Negative Logits

 kaynağından

-0.42

googleapis

-0.42

"]').

-0.40

bootstrapcdn

-0.40

ListItemIcon

-0.40

 للمعارف

-0.40

 MainAxisSize

-0.40

ArgsConstructor

-0.40

IsMutable

-0.39

Jereo

-0.39

POSITIVE LOGITS

 فريبيس

0.34

kriv

0.30

 Sitz

0.30

uyt

0.28

 Seil

0.28

]--;

0.27

 ſtate

0.27

 observations

0.27

 observaciones

0.26

ApiResponse

0.26

Activations Density 0.001%

attends to tokens related to the expressions and commands from programming or technical contexts

No Comments

No Known Activations