INDEX

Explanations

attends to figure-related tokens from subsequent relevant content

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

0:0.01

1:0.03

2:0.05

3:0.20

4:0.04

5:0.06

6:0.49

7:0.08

Negative Logits

 itſelf

-0.47

 myſelf

-0.47

 فريبيس

-0.46

\{\\

-0.46

IsContent

-0.46

 themſelves

-0.46

ſelves

-0.45

 EconPapers

-0.45

 különböz

-0.45

 ddelweddau

-0.44

POSITIVE LOGITS

MigrationBuilder

0.39

''}

0.32

saida

0.31

 usos

0.31

Климат

0.31

ról

0.30

--;

0.30

Esprit

0.29

/−

0.29

'}>

0.29

Activations Density 0.862%

attends to figure-related tokens from subsequent relevant content

No Comments

No Known Activations