INDEX

Explanations

attends to token patterns that contain specific keywords or phrases in the format capturing sub-entities or attributes representative of the corresponding sequential nature

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.04

1:0.09

2:0.05

3:0.04

4:0.21

5:0.45

6:0.04

7:0.04

Negative Logits

MemoryWarning

-0.33

Eloquent

-0.28

 UIFont

-0.26

HomeAsUpEnabled

-0.26

jScrollPane

-0.26

FlatList

-0.26

 sanitaria

-0.25

 jLabel

-0.25

SequentialGroup

-0.25

ьаж

-0.24

POSITIVE LOGITS

DebuggerNonUser

0.35

contentLoaded

0.30

thington

0.29

لكتر

0.29

Batis

0.28

ceptives

0.28

 Exactos

0.28

principalColumn

0.27

XmlAccessType

0.27

ărilor

0.27

Activations Density 0.924%

attends to token patterns that contain specific keywords or phrases in the format capturing sub-entities or attributes representative of the corresponding sequential nature

No Comments

No Known Activations