INDEX

Explanations

mentions of various groups of people and their activities or interests

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_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.3.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

0:0.01

1:0.05

2:0.08

3:0.02

4:0.01

5:0.05

6:0.05

7:0.05

8:0.04

9:0.48

10:0.04

11:0.06

Negative Logits

 forward

-1.35

 envelope

-1.32

egu

-1.26

bow

-1.24

 suite

-1.24

tty

-1.23

 squad

-1.22

 alive

-1.21

 srfAttach

-1.21

 slate

-1.20

POSITIVE LOGITS

 sake

2.12

icion

1.64

 purposes

1.58

pired

1.47

gencies

1.33

 Psychic

1.32

protection

1.32

umption

1.31

recated

1.30

ulz

1.26

Activations Density 0.229%

mentions of various groups of people and their activities or interests

No Comments

No Known Activations