INDEX

Explanations

attends to specific nouns or concepts from related earlier tokens in the sequence

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

0:0.10

1:0.08

2:0.09

3:0.40

4:0.08

5:0.03

6:0.08

7:0.10

Negative Logits

openzeppelin

-0.42

ChildScrollView

-0.38

visiae

-0.37

WEBPACK

-0.35

gpl

-0.35

 ujednoznacz

-0.35

([[

-0.35

gehend

-0.34

ündig

-0.34

])[

-0.33

POSITIVE LOGITS

 '\\;'

0.45

 HasFactory

0.40

 EconPapers

0.36

 jsPsych

0.34

fitrión

0.33

 kasarigan

0.33

 विश्वसनीयता

0.32

 Wikimedijinoj

0.32

HasAnnotation

0.31

Activations Density 0.670%

attends to specific nouns or concepts from related earlier tokens in the sequence

No Comments

No Known Activations