INDEX

Explanations

attends to compiler-related tokens from various programming or technical 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.02

1:0.03

2:0.12

3:0.45

4:0.08

5:0.08

6:0.09

7:0.09

Negative Logits

-0.30

 pleaſure

-0.27

 purpoſe

-0.27

-0.25

-0.24

-0.23

POSITIVE LOGITS

UnusedPrivate

0.58

 propOrder

0.55

AddTagHelper

0.55

rungsseite

0.54

NameInMap

0.54

 Roskov

0.54

SequentialGroup

0.53

>');

0.53

="@+

0.51

DebuggerNonUser

0.50

Activations Density 0.320%

attends to compiler-related tokens from various programming or technical tokens

No Comments

No Known Activations