INDEX

Explanations

attends to the token "2" from various mathematical expressions or sequences containing numbers

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

0:0.28

1:0.25

2:0.12

3:0.06

4:0.08

5:0.04

6:0.04

7:0.08

Negative Logits

es

-0.26

δ

-0.25

ke

-0.25

Ch

-0.25

is

-0.24

UnitTesting

-0.24

ia

-0.23

AB

-0.23

 estimés

-0.23

 fact

-0.23

POSITIVE LOGITS

cautionary

0.32

NSIndexPath

0.31

 occaf

0.30

 Савезне

0.30

 fevere

0.30

 dreamstime

0.30

 auffi

0.30

 ExecuteAsync

0.30

 uſed

0.30

✨:

0.29

Activations Density 1.752%

attends to the token "2" from various mathematical expressions or sequences containing numbers

No Comments

No Known Activations