INDEX

Explanations

the word "only."

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

0:0.08

1:0.07

2:0.07

3:0.08

4:0.07

5:0.07

6:0.08

7:0.08

8:0.08

9:0.08

10:0.08

11:0.09

Negative Logits

adobe

-2.81

Gamer

-2.63

 Moto

-2.56

owa

-2.43

gpu

-2.39

 Bunny

-2.39

INO

-2.38

Ko

-2.37

}}}

-2.36

gamer

-2.30

POSITIVE LOGITS

 instr

2.56

alion

2.51

 ministry

2.40

 apparatus

2.38

 recruits

2.32

lex

2.30

 Hess

2.25

 recruitment

2.24

loc

2.23

 curric

2.20

Activations Density 0.000%

the word "only."

No Comments

No Known Activations