INDEX

Explanations

attends to specific tokens marked for detailed explanations from the broader context of scientific studies or methodology

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

0:0.11

1:0.13

2:0.09

3:0.13

4:0.12

5:0.06

6:0.14

7:0.18

Negative Logits

bootstrapcdn

-0.35

 defaultstate

-0.34

rrggbb

-0.34

 tumours

-0.32

schaft

-0.31

 propOrder

-0.31

 Prieto

-0.30

 tbsp

-0.30

IBOutlet

-0.30

SpringBootTest

-0.30

POSITIVE LOGITS

 referrerpolicy

0.36

 الاطلاع

0.35

 समीक्षाओं

0.31

Slf

0.31

findpost

0.30

—

0.30

Geplaatst

0.29

eace

0.29

 PARSER

0.28

PAD

0.28

Activations Density 0.002%

attends to specific tokens marked for detailed explanations from the broader context of scientific studies or methodology

No Comments

No Known Activations