INDEX

Explanations

attends to various prepositions and conjunctions that connect phrases and ideas from tokens related to scientific or technical contexts

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

0:0.08

1:0.16

2:0.42

3:0.06

4:0.06

5:0.03

6:0.05

7:0.09

Negative Logits

elemField

-0.41

Datuak

-0.38

tagHelperRunner

-0.36

MLLoader

-0.35

 Infórmanos

-0.35

TintMode

-0.34

 שוליים

-0.33

 cherchés

-0.33

>*/

-0.33

POSITIVE LOGITS

 satunya

0.27

 vertrouwen

0.26

ordning

0.25

explique

0.24

dout

0.24

AMIENTO

0.24

Daughter

0.23

 aikaa

0.23

 hosting

0.23

לים

0.23

Activations Density 1.134%

attends to various prepositions and conjunctions that connect phrases and ideas from tokens related to scientific or technical contexts

No Comments

No Known Activations