INDEX

Explanations

attends to terms related to annual measures from terms that indicate totals

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

0:0.20

1:0.21

2:0.13

3:0.10

4:0.09

5:0.03

6:0.08

7:0.13

Negative Logits

bootstrapcdn

-0.42

 GenerationType

-0.38

يميديا

-0.38

harusnya

-0.37

SequentialGroup

-0.37

 Humphries

-0.36

FDP

-0.35

minecraft

-0.35

BNF

-0.35

 ddelweddau

-0.35

POSITIVE LOGITS

 الاطلاع

0.36

 Ballard

0.35

aik

0.33

 Walling

0.33

garis

0.31

ellis

0.30

 Lehr

0.29

 Trost

0.29

 Honors

0.29

leta

0.29

Activations Density 8.880%

attends to terms related to annual measures from terms that indicate totals

No Comments

No Known Activations