INDEX

Explanations

attends to tokens representing parameters or arguments from tokens indicating function calls or data inputs

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

0:0.09

1:0.09

2:0.50

3:0.06

4:0.06

5:0.04

6:0.04

7:0.07

Negative Logits

BeginInit

-0.48

 estimés

-0.38

///</

-0.37

PyExc

-0.37

parsedMessage

-0.37

 مرئيه

-0.36

참고

-0.35

 AppCompat

-0.35

 distanciation

-0.35

//</

-0.34

POSITIVE LOGITS

ftagPool

0.26

 gyhoeddwyd

0.25

 dead

0.25

 apes

0.23

стероид

0.23

ZoneId

0.23

 VIVO

0.23

habad

0.22

 soldiers

0.22

 Clive

0.22

Activations Density 0.335%

attends to tokens representing parameters or arguments from tokens indicating function calls or data inputs

No Comments

No Known Activations