INDEX

Explanations

attends to the token "printf" from programming-related 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.01

1:0.03

2:0.06

3:0.55

4:0.03

5:0.06

6:0.14

7:0.08

Negative Logits

numerusform

-0.53

StandardCharsets

-0.42

存于互联网档案馆

-0.41

}\]

-0.41

AndEndTag

-0.41

AddTagHelper

-0.41

 ExecuteAsync

-0.40

itoneal

-0.40

SequentialGroup

-0.40

>';

-0.40

POSITIVE LOGITS

 giuri

0.29

 ſta

0.26

 sandero

0.26

Efq

0.25

etc

0.24

 juſt

0.24

 ejus

0.24

 espagne

0.23

ed

0.23

 karde

0.23

Activations Density 0.104%

attends to the token "printf" from programming-related tokens

No Comments

No Known Activations