INDEX

Explanations

attends to the token "on" from instances of the same token in the sequence as well as from other contextual 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

0:0.02

1:0.18

2:0.46

3:0.02

4:0.04

5:0.03

6:0.03

7:0.19

Negative Logits

on

-0.57

in

-0.52

an

-0.36

 deschis

-0.35

and

-0.35

 Cæsar

-0.35

 umano

-0.35

 morire

-0.34

 subterrá

-0.33

to

-0.33

POSITIVE LOGITS

tagHelperRunner

0.73

rungsseite

0.65

sizeCache

0.63

 propOrder

0.61

TypedDataSet

0.61

aarrggbb

0.61

findpost

0.60

ⓧ

0.59

WithIOException

0.59

ThroughAttribute

0.59

Activations Density 0.045%

attends to the token "on" from instances of the same token in the sequence as well as from other contextual tokens

No Comments

No Known Activations