INDEX

Explanations

attends to empty or boundary markers from arbitrary 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

0:0.06

1:0.16

2:0.08

3:0.07

4:0.25

5:0.19

6:0.08

7:0.08

Negative Logits

 itſelf

-3.95

 myſelf

-3.91

Efq

-3.77

 Monfieur

-3.56

 Jefus

-3.55

 Theſe

-3.45

 Majefty

-3.39

 ―――――

-3.39

 pleaſure

-3.39

ſelf

-3.31

POSITIVE LOGITS

2.03

↵

2.00

1.94

<eos>

1.93

↵↵

1.91

1.74

1.70

in

1.67

Activations Density 1.052%

attends to empty or boundary markers from arbitrary subsequent tokens

No Comments

No Known Activations