INDEX

Explanations

attends to tokens with the "ll" pattern from subsequent tokens

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

0:0.82

1:0.01

2:0.01

3:0.01

4:0.07

5:0.02

6:0.01

7:0.02

Negative Logits

get

-0.61

be

-0.57

 have

-0.56

 find

-0.53

see

-0.53

 think

-0.52

 consider

-0.51

 make

-0.49

 take

-0.49

 understand

-0.47

POSITIVE LOGITS

SourceChecksum

0.47

OGND

0.46

RegressionTest

0.45

ьаж

0.39

izr

0.39

 veckor

0.39

ensement

0.38

</caption>

0.37

 }));

0.37

 Galile

0.36

Activations Density 0.026%

attends to tokens with the "ll" pattern from subsequent tokens

No Comments

No Known Activations