INDEX

Explanations

attends to various assertions about specific terms from subsequent references to those terms

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

0:0.14

1:0.22

2:0.19

3:0.08

4:0.10

5:0.06

6:0.08

7:0.08

Negative Logits

MessageOf

-0.46

期刊论文

-0.35

principalTable

-0.34

__':

-0.33

__((

-0.32

fromnode

-0.32

tonode

-0.31

 Вікіпе

-0.31

"]').

-0.30

phazard

-0.30

POSITIVE LOGITS

 برانيه

0.33

 numberWith

0.33

 jaya

0.33

Portale

0.33



0.32

urlpatterns

0.32

 externi

0.32

cipit

0.32

thema

0.32

Screens

0.32

Activations Density 0.165%

attends to various assertions about specific terms from subsequent references to those terms

No Comments

No Known Activations