INDEX

Explanations

athlete names

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.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

0:0.11

1:0.04

2:0.09

3:0.07

4:0.05

5:0.12

6:0.07

7:0.05

8:0.10

9:0.09

10:0.12

11:0.05

Negative Logits

andel

-1.47

eros

-1.27

龍�

-1.25

龍喚士

-1.25

ecause

-1.19

Santa

-1.17

manent

-1.14

Different

-1.14

Demand

-1.13

aea

-1.12

POSITIVE LOGITS

 Canaver

1.38

 meanwhile

1.31

 moreover

1.27

 Spiel

1.25

 spokesman

1.23

 Bris

1.21

Osw

1.20

Yus

1.19

 ital

1.19

 Rosenthal

1.18

Activations Density 0.125%

athlete names

No Comments

No Known Activations