INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.07

1:0.13

2:0.08

3:0.07

4:0.07

5:0.06

6:0.10

7:0.08

8:0.08

9:0.07

10:0.08

11:0.08

Negative Logits

QC

-1.69

 Interior

-1.66

 Levin

-1.61

 Maximum

-1.59

 Possible

-1.53

 Permanent

-1.52

 Legacy

-1.48

wreck

-1.48

 Bleach

-1.47

 mend

-1.47

POSITIVE LOGITS

inav

2.18

tera

1.79

onom

1.78

 bases

1.73

yip

1.72

1.67

hift

1.63

 carts

1.62

ベ

1.61

 spons

1.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.