INDEX

Explanations

names or references to specific individuals

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

0:0.07

1:0.04

2:0.22

3:0.08

4:0.16

5:0.06

6:0.03

7:0.03

8:0.07

9:0.09

10:0.06

11:0.02

Negative Logits

 ponies

-1.24

ascript

-1.09

ixir

-1.08

kefeller

-1.08

atron

-1.07

 loaded

-1.06

 ancest

-1.06

 equals

-1.05

 Retrieved

-1.02

 unloaded

-1.02

POSITIVE LOGITS

pta

1.26

Khe

1.26

gha

1.24

oso

1.24

ni

1.23

nen

1.22

insk

1.21

カ

1.20

orst

1.18

aan

1.18

Activations Density 0.002%

names or references to specific individuals

No Comments

No Known Activations