INDEX

Explanations

attends to the token "sub" from various terms that follow with related second tokens

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

0:0.10

1:0.02

2:0.01

3:0.02

4:0.73

5:0.03

6:0.02

7:0.04

Negative Logits

 fubject

-0.47

 Integr

-0.45

 theſe

-0.44

 poffible

-0.44

 fevere

-0.43

 purpoſe

-0.42

Efq

-0.42

Poz

-0.42

rizona

-0.42

 leaſt

-0.41

POSITIVE LOGITS

sub

0.36

 OFDb

0.34

 дописавши

0.33

 EconPapers

0.33

 Chwiliwch

0.32

ได้

0.32

He

0.31

 also

0.31

yntaxException

0.31

ilerini

0.31

Activations Density 0.076%

attends to the token "sub" from various terms that follow with related second tokens

No Comments

No Known Activations