INDEX

Explanations

attends to formatted scientific variables and units from generic descriptors that indicate various formats of measurement or notation

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

1:0.11

2:0.05

3:0.04

4:0.13

5:0.51

6:0.04

7:0.04

Negative Logits

RegressionTest

-0.57

 Réponses

-0.55

parsedMessage

-0.53

Vidite

-0.51

Tikang

-0.50

addCriterion

-0.50

 &___

-0.50

RTDA

-0.46

GEBURTSDATUM

-0.45

 مرئيه

-0.45

POSITIVE LOGITS

श्व

0.37

AndEndTag

0.36

plak

0.34

0.32

 vigueur

0.31

󠁮

0.31

by

0.30

 فريبيس

0.30

󠁢

0.30

 Venta

0.29

Activations Density 0.165%

attends to formatted scientific variables and units from generic descriptors that indicate various formats of measurement or notation

No Comments

No Known Activations