INDEX

Explanations

phrases indicating participation or inclusion in activities or events

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

0:0.02

1:0.02

2:0.06

3:0.06

4:0.11

5:0.02

6:0.03

7:0.45

8:0.03

9:0.03

10:0.07

11:0.05

Negative Logits

gency

-1.67

amera

-1.44

inity

-1.42

��

-1.34

 pharmacies

-1.31

uds

-1.31

obar

-1.30

ges

-1.30

ifax

-1.28

 sued

-1.27

POSITIVE LOGITS

 bandwagon

1.64

 festivities

1.58

 proceedings

1.45

 ratification

1.44

 Unch

1.40

 routines

1.39

feat

1.38

 fray

1.38

accompan

1.38

Cla

1.36

Activations Density 0.007%

phrases indicating participation or inclusion in activities or events

No Comments

No Known Activations