INDEX

Explanations

attends to tokens indicating a clarifying or defining relationship in the context of statements from tokens providing the respective statements and their summaries

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

0:0.07

1:0.07

2:0.08

3:0.10

4:0.07

5:0.02

6:0.33

7:0.23

Negative Logits

__(/*!

-0.25

แท

-0.25

viewDidLoad

-0.25

 TextInputType

-0.25

findpost

-0.24

“

-0.24

vacy

-0.24

ano

-0.23

all

-0.23

CodeAnalysis

-0.23

POSITIVE LOGITS

 Roskov

0.38

 iprot

0.32

 dieux

0.32

 cavalo

0.32

 Belk

0.31

 EnglishChoose

0.30

hopefully

0.30

:✨

0.30

gradable

0.30

 zufolge

0.29

Activations Density 0.518%

attends to tokens indicating a clarifying or defining relationship in the context of statements from tokens providing the respective statements and their summaries

No Comments

No Known Activations