INDEX

Explanations

attends to specific numerical significance tokens from various results

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

0:0.13

1:0.22

2:0.11

3:0.05

4:0.08

5:0.10

6:0.06

7:0.22

Negative Logits

BeanUtils

-0.35

providedIn

-0.33

chenkt

-0.31

 BioLib

-0.31

ResumeLayout

-0.30

WriteAttribute

-0.30

parseColor

-0.29

 Roost

-0.28

 VIAF

-0.28

WebpackPlugin

-0.28

POSITIVE LOGITS

Efq

0.43

ỡng

0.43

 whoſe

0.42

/>";

0.39

NUMX

0.39

.";

0.39

//

0.39

;";

0.39

 disambiguazione

0.39

/**

0.38

Activations Density 0.388%

attends to specific numerical significance tokens from various results

No Comments

No Known Activations