INDEX

Explanations

phrases related to safety and security

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

0:0.02

1:0.02

2:0.05

3:0.04

4:0.13

5:0.03

6:0.41

7:0.06

8:0.02

9:0.04

10:0.07

11:0.05

Negative Logits

ottest

-1.41

NL

-1.30

atform

-1.28

agher

-1.22

neum

-1.21

ovych

-1.21

VICE

-1.19

ften

-1.19

gem

-1.18

###

-1.15

POSITIVE LOGITS

 clauses

1.31

 PowerPoint

1.25

 pronouns

1.25

 notation

1.25

 manual

1.22

utions

1.21

 Quote

1.20

 discretion

1.18

 gestures

1.17

 noun

1.16

Activations Density 0.035%

phrases related to safety and security

No Comments

No Known Activations