INDEX

Explanations

the presence of parentheses in the text

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.9.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.09

1:0.07

2:0.08

3:0.08

4:0.09

5:0.08

6:0.08

7:0.07

8:0.07

9:0.09

10:0.07

11:0.08

Negative Logits

 paternity

-3.30

breeding

-3.29

 Helena

-2.77

 Elias

-2.72

 adam

-2.59

 prophes

-2.55

 mating

-2.54

 Huma

-2.54

 Isabel

-2.53

 indoctr

-2.53

POSITIVE LOGITS

ゴン

3.60

inburgh

3.31

rawdownloadcloneembedreportprint

3.08

KEN

2.98

AME

2.96

ヤ

2.84

neapolis

2.77

Export

2.77

ロ

2.77

soDeliveryDate

2.75

Activations Density 0.000%

the presence of parentheses in the text

No Known Activations