INDEX

Explanations

phrases indicating a point of origin or source

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

1:0.01

2:0.09

3:0.07

4:0.20

5:0.02

6:0.07

7:0.25

8:0.03

9:0.02

10:0.08

11:0.08

Negative Logits

payer

-1.56

apsed

-1.42

eday

-1.36

 twent

-1.34

 pled

-1.33

 tubing

-1.32

chini

-1.30

nz

-1.27

 subscrib

-1.27

bryce

-1.26

POSITIVE LOGITS

capital

1.65

 distraction

1.57

mist

1.52

Stock

1.51

erno

1.48

 sins

1.46

 neglect

1.45

 blows

1.44

Winter

1.39

balance

1.39

Activations Density 0.004%

phrases indicating a point of origin or source

No Comments

No Known Activations