INDEX

Explanations

attends to specific tokens linked to antagonistic contexts from related supportive 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

0:0.09

1:0.29

2:0.17

3:0.05

4:0.13

5:0.06

6:0.05

7:0.13

Negative Logits

 >=",

-0.36

 disponibilités

-0.35

wyt

-0.34

oa̍t

-0.34

expandindo

-0.34

GMT

-0.33

坞

-0.33

enciaga

-0.33

 Propriété

-0.32

 Autorizaciones

-0.32

POSITIVE LOGITS

compañ

0.35

 initComponents

0.33

dė

0.33

UrlResolution

0.29

 récon

0.27

lepší

0.26

Билгалдахарш

0.26

abestanden

0.26

reciate

0.26

 Gutes

0.26

Activations Density 0.609%

attends to specific tokens linked to antagonistic contexts from related supportive tokens

No Comments

No Known Activations