INDEX

Explanations

attends to the various instances of tokens that are modified by their own duplicate forms

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

0:0.05

1:0.09

2:0.05

3:0.07

4:0.50

5:0.11

6:0.04

7:0.05

Negative Logits

__*/

-0.34

__':

-0.28



-0.28

 становника

-0.27

 isKindOfClass

-0.27

findpost

-0.27

کر

-0.26

日閲覧

-0.26

 consultato

-0.26

 JADX

-0.26

POSITIVE LOGITS

prüche

0.36

expandindo

0.31

InstrumentedTest

0.30

xicity

0.28

niczy

0.28

هما

0.28

gemini

0.28

hastly

0.27

 Ashraf

0.26

estacks

0.26

Activations Density 1.429%

attends to the various instances of tokens that are modified by their own duplicate forms

No Comments

No Known Activations