INDEX

Explanations

attends to implications or consequences marked by "to" from introductory or conditional phrases marked by "In."

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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.13

1:0.17

2:0.11

3:0.08

4:0.05

5:0.02

6:0.09

7:0.32

Negative Logits

kir

-0.27

 viſ

-0.27

 faſt

-0.25

 uſe

-0.24

 anſ

-0.24

 vallée

-0.24

NAMES

-0.23

 purpoſe

-0.23

 reaſon

-0.23

 ſhall

-0.23

POSITIVE LOGITS

脚注の使い方

0.39

AnimationsModule

0.35

tvguidetime

0.33

 Bahamas

0.31

apimachinery

0.30

istoitu

0.30

getDoctrine

0.30

NgModule

0.30

Massa

0.30

JsonProperty

0.28

Activations Density 0.053%

attends to implications or consequences marked by "to" from introductory or conditional phrases marked by "In."

No Comments

No Known Activations