INDEX

Explanations

attends to specific variables and functions denoted by parentheses and brackets, focusing on their parameters and structures from a preceding context

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

0:0.23

1:0.23

2:0.14

3:0.11

4:0.07

5:0.04

6:0.04

7:0.10

Negative Logits

RenderAtEndOf

-0.33

 gynnwys

-0.29

Gim

-0.27

 "..\..\..\

-0.27

minder

-0.26

mtrl

-0.25

клопе

-0.25

 Preferencias

-0.25

Begriffsklä

-0.24

TemporalType

-0.24

POSITIVE LOGITS

didSet

0.32

ambién

0.28

pshots

0.28

 financiera

0.27

 ویکی‌پدیا

0.27

numerusform

0.27

 scienza

0.26

 Suara

0.26

جوايز

0.25

optString

0.25

Activations Density 0.548%

attends to specific variables and functions denoted by parentheses and brackets, focusing on their parameters and structures from a preceding context

No Comments

No Known Activations