INDEX

Explanations

locations and positions associated with individuals or groups in a context

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_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

1:0.01

2:0.07

3:0.05

4:0.09

5:0.01

6:0.04

7:0.46

8:0.02

9:0.02

10:0.09

11:0.08

Negative Logits

chieve

-1.64

レ

-1.63

aan

-1.63

breaks

-1.60

 injure

-1.55

 encount

-1.51

 eradicate

-1.49

infect

-1.48

ェ

-1.48

ット

-1.46

POSITIVE LOGITS

 Mouse

1.75

 wheel

1.57

 chairs

1.56

 Tsukuyomi

1.50

 Throne

1.49

 Cinderella

1.47

 Garr

1.43

RIS

1.42

 throne

1.41

 Circuit

1.41

Activations Density 0.011%

locations and positions associated with individuals or groups in a context

No Comments

No Known Activations