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

0:0.07

1:0.08

2:0.09

3:0.07

4:0.09

5:0.08

6:0.08

7:0.07

8:0.08

9:0.09

10:0.08

11:0.06

Negative Logits

 Crime

-3.24

 Challenge

-3.00

 Graves

-2.92

 Lifetime

-2.71

 Household

-2.68

 Detective

-2.61

 Crimes

-2.60

 Passage

-2.57

 Trophy

-2.57

 Guerrero

-2.57

POSITIVE LOGITS

Sov

3.14

 laun

2.95

liber

2.83

_>

2.70

ublic

2.58

kernel

2.56

Layout

2.51

Liber

2.48

hawks

2.43

Cre

2.42

Activations Density 0.000%

the word "only."

No Comments

No Known Activations