INDEX

Explanations

attends to the emphasis or confirmation tokens from clarifying questions or responses

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

1:0.10

2:0.10

3:0.12

4:0.10

5:0.04

6:0.23

7:0.21

Negative Logits

PopupMenu

-0.26

board

-0.24

zas

-0.24

 بما

-0.23

<eos>

-0.23

Week

-0.22

boards

-0.22

thra

-0.22

 proves

-0.22

Dob

-0.21

POSITIVE LOGITS

 jspb

0.42

DockStyle

0.40

 المعيارى

0.39

 myſelf

0.39

setVerticalGroup

0.39

Rhestr

0.37

 nôtre

0.37

 precisione

0.36

 quelconque

0.36

awtextra

0.36

Activations Density 0.037%

attends to the emphasis or confirmation tokens from clarifying questions or responses

No Comments

No Known Activations