INDEX

Explanations

attends to keywords and punctuation marks from tokens related to academic discussions and literature

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

0:0.08

1:0.11

2:0.29

3:0.24

4:0.10

5:0.04

6:0.05

7:0.06

Negative Logits

 EconPapers

-0.33

JAKARTA

-0.32

 légales

-0.32

tvguidetime

-0.30

ille

-0.28

AKI

-0.28

 quil

-0.28

zak

-0.28

å

-0.27

ടെ

-0.27

POSITIVE LOGITS

 nakalista

0.39

RUnlock

0.39

 resourceCulture

0.39

 wireType

0.37

PerformLayout

0.35

extAlignment

0.32

ConstraintMaker

0.32

Parcelize

0.32

ьаж

0.31

encodeWith

0.31

Activations Density 0.098%

attends to keywords and punctuation marks from tokens related to academic discussions and literature

No Comments

No Known Activations