INDEX

Explanations

information related to names and personal identifiers

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_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

0:0.06

1:0.03

2:0.03

3:0.07

4:0.03

5:0.04

6:0.15

7:0.03

8:0.03

9:0.40

10:0.03

11:0.03

Negative Logits

ple

-3.99

 Thom

-3.95

Vog

-3.89

Rom

-3.77

 Noble

-3.66

vae

-3.39

 Romans

-3.38

 obser

-3.30

 Flore

-3.24

Vis

-3.21

POSITIVE LOGITS

ACK

6.58

ack

6.55

acks

6.30

rack

5.46

 Mack

5.46

ackle

5.39

acking

5.28

 Shack

5.28

acker

5.12

acked

5.04

Activations Density 0.001%

information related to names and personal identifiers

No Comments

No Known Activations