INDEX

Explanations

phrases indicating decision-making and choices

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

0:0.01

1:0.01

2:0.14

3:0.15

4:0.11

5:0.02

6:0.03

7:0.29

8:0.04

9:0.03

10:0.08

11:0.03

Negative Logits

gif

-1.65

ingo

-1.64

Toy

-1.60

キ

-1.50

sound

-1.47

ェ

-1.46

objects

-1.44

interstitial

-1.42

agram

-1.41

video

-1.39

POSITIVE LOGITS

 calibr

1.88

 discretion

1.75

 intervene

1.75

 willingness

1.72

 contingency

1.72

 certify

1.72

 manpower

1.72

 patience

1.71

 decide

1.70

 suscept

1.66

Activations Density 0.004%

phrases indicating decision-making and choices

No Known Activations