INDEX

Explanations

attends to "similar" from various related tokens indicating comparison or analogy

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

0:0.67

1:0.01

2:0.01

3:0.01

4:0.18

5:0.03

6:0.01

7:0.05

Negative Logits

 itſelf

-0.40

 Shakspeare

-0.35

 ſche

-0.34

 ſta

-0.33

 purpoſe

-0.33

 ſch

-0.33

 myſelf

-0.32

 Hift

-0.32

 Diſ

-0.31

{},

-0.31

POSITIVE LOGITS

GHIJKLM

0.50

providedIn

0.46

Pautan

0.46

 ostavi

0.45

 tartalomajánló

0.42

 linkovi

0.41

 propOrder

0.40

 pinulongan

0.40

 actionTypes

0.39

 gynhyrchwyd

0.39

Activations Density 0.036%

attends to "similar" from various related tokens indicating comparison or analogy

No Comments

No Known Activations