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

Head Attr Weights

0:0.09

1:0.05

2:0.09

3:0.08

4:0.08

5:0.08

6:0.06

7:0.11

8:0.08

9:0.08

10:0.07

11:0.09

Negative Logits

 Penguin

-1.96

orsche

-1.73

SAS

-1.66

 unwanted

-1.62

seless

-1.62

 Romeo

-1.60

 IMAGES

-1.60

 offending

-1.60

 infringing

-1.55

 rodent

-1.53

POSITIVE LOGITS

gemony

2.14

)?

1.81

vati

1.78

wed

1.65

zinski

1.60

ismo

1.53

umen

1.52

etc

1.52

AMA

1.51

allowed

1.51

Activations Density 0.000%

No Known Activations

This feature has no known activations.