INDEX

Explanations

attends to tokens related to function calls or method invocations from specific structured data or records

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.01

1:0.03

2:0.63

3:0.08

4:0.05

5:0.05

6:0.07

7:0.04

Negative Logits

 myſelf

-1.39

 itſelf

-1.36

Efq

-1.31

 Monfieur

-1.25

 ―――――

-1.23

 Theſe

-1.20

 Majefty

-1.20

 pleaſure

-1.18

 Jefus

-1.18

 auffi

-1.18

POSITIVE LOGITS

0.80

0.79

0.75

<eos>

0.73

0.70

↵↵

0.68

in

0.64

0.63

0.62

Activations Density 0.199%

attends to tokens related to function calls or method invocations from specific structured data or records

No Comments

No Known Activations