INDEX

Explanations

statements or phrases that indicate conclusions

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_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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.05

1:0.02

2:0.11

3:0.05

4:0.10

5:0.09

6:0.04

7:0.03

8:0.10

9:0.24

10:0.07

11:0.04

Negative Logits

ustomed

-1.30

cel

-1.27

unts

-1.21

Joined

-1.20

apon

-1.17

 broom

-1.11

bill

-1.07

lav

-1.07

 abroad

-1.06

 brushes

-1.06

POSITIVE LOGITS

��

1.46

���

1.42

 Forth

1.30

 captcha

1.27

 explan

1.24

��

1.24

 causation

1.23

 moot

1.23

 disapp

1.21

 Preferences

1.21

Activations Density 0.014%

statements or phrases that indicate conclusions

No Comments

No Known Activations