INDEX

Explanations

phrases related to safety and security

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

0:0.04

1:0.03

2:0.22

3:0.05

4:0.16

5:0.08

6:0.02

7:0.03

8:0.06

9:0.19

10:0.05

11:0.03

Negative Logits

itcher

-1.61

secut

-1.43

cin

-1.39

aucas

-1.37

JB

-1.30

laugh

-1.29

incial

-1.29

ench

-1.27

oples

-1.27

 apiece

-1.25

POSITIVE LOGITS

 confidentiality

1.44

Safety

1.40

osate

1.39

tnc

1.38

ForgeModLoader

1.38

 safety

1.35

HIP

1.35

 protections

1.32

 safeguards

1.29

 precautions

1.29

Activations Density 0.028%

phrases related to safety and security

No Comments

No Known Activations