INDEX

Explanations

references to specific people, possibly related to politics or current events

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.10

1:0.03

2:0.26

3:0.08

4:0.12

5:0.05

6:0.03

7:0.03

8:0.04

9:0.14

10:0.05

11:0.02

Negative Logits

ngth

-1.67

 torches

-1.50

yip

-1.37

 Illum

-1.30

itiveness

-1.30

etheless

-1.29

emonium

-1.27

holders

-1.27

オ

-1.25

ificant

-1.24

POSITIVE LOGITS

lez

1.31

ene

1.27

lish

1.20

 herself

1.19

lene

1.15

etus

1.14

zyk

1.14

inski

1.14

sey

1.13

tree

1.12

Activations Density 0.010%

references to specific people, possibly related to politics or current events

No Comments

No Known Activations