INDEX

Explanations

attends to the dark areas or themes from various contexts related to tokens may be flexible

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

0:0.74

1:0.01

2:0.01

3:0.02

4:0.11

5:0.02

6:0.01

7:0.04

Negative Logits

出版年

-0.38

 propOrder

-0.37

AddTagHelper

-0.37

chymal

-0.33

 Signalez

-0.33

NOPQRST

-0.32

enterOuterAlt

-0.31

UnsafeEnabled

-0.31

homonymie

-0.30

 protoimpl

-0.30

POSITIVE LOGITS

.";

0.32

 reguli

0.31

tür

0.31

ttino

0.31

kloped

0.31

'',

0.30

**/

0.30

oucí

0.30

cency

0.30

'';

0.30

Activations Density 0.007%

attends to the dark areas or themes from various contexts related to tokens may be flexible

No Comments

No Known Activations