INDEX

Explanations

attends to specific names and terms, such as a name or role, from various related tokens

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

1:0.14

2:0.43

3:0.05

4:0.05

5:0.02

6:0.04

7:0.09

Negative Logits

ing

-0.30

ligen

-0.28

Cyfeiriadau

-0.26

ราย

-0.26

opp

-0.26

HasIndex

-0.26

нибудь

-0.25

brz

-0.25

wan

-0.25

any

-0.24

POSITIVE LOGITS

 Мексичка

0.38

 becauſe

0.36

sidemargin

0.35

Portale

0.35

 myſelf

0.34

 ModelRenderer

0.33

 poffible

0.33

Efq

0.33

+#+

0.33

 iſt

0.31

Activations Density 2.048%

attends to specific names and terms, such as a name or role, from various related tokens

No Comments

No Known Activations