INDEX

Explanations

phrases indicating contrasts or variations in experiences

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.03

1:0.20

2:0.04

3:0.04

4:0.07

5:0.02

6:0.04

7:0.17

8:0.03

9:0.04

10:0.06

11:0.21

Negative Logits

 pseudo

-2.60

Effective

-2.55

 obsolete

-2.52

 aquatic

-2.51

 srfAttach

-2.50

 adjunct

-2.46

 Survive

-2.40

 preschool

-2.40

ZE

-2.39

effective

-2.35

POSITIVE LOGITS

 downs

3.34

ups

3.16

 Downs

2.84

 nerves

2.83

 bumps

2.83

Mania

2.65

 Priebus

2.65

 highs

2.58

dinand

2.57

 occasions

2.54

Activations Density 0.000%

phrases indicating contrasts or variations in experiences

No Known Activations