INDEX

Explanations

attends to type definitions from corresponding tokens that denote specific attributes or categories

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

1:0.11

2:0.11

3:0.07

4:0.06

5:0.03

6:0.12

7:0.38

Negative Logits

Према

-0.29

osoba

-0.29

StoryboardSegue

-0.28

⚧

-0.27

ilber

-0.27

Embeddable

-0.26

 Waray

-0.26

 ISNI

-0.25

 trans

-0.25

 troppo

-0.25

POSITIVE LOGITS

 становника

0.35

 iyon

0.34

stateProvider

0.33

Efq

0.32

fromLTRB

0.32

enumi

0.32

ToTensor

0.31

 itſelf

0.31

 loob

0.31

InjectAttribute

0.31

Activations Density 0.069%

attends to type definitions from corresponding tokens that denote specific attributes or categories

No Comments

No Known Activations