INDEX

Explanations

phrases indicating conditional reasoning or causation

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.02

1:0.00

2:0.27

3:0.17

4:0.08

5:0.03

6:0.07

7:0.07

8:0.05

9:0.04

10:0.06

11:0.09

Negative Logits

sid

-1.45

flush

-1.37

 favourites

-1.31

 plates

-1.26

ength

-1.25

 favorites

-1.23

 COURT

-1.23

 Onion

-1.21

 Elsa

-1.20

ullivan

-1.19

POSITIVE LOGITS

◼

1.78

��

1.74

��

1.71

��

1.46

 persistence

1.42

 ********************************

1.38

$$$$

1.35

 Archdemon

1.35

regation

1.33

 conspicuous

1.32

Activations Density 0.004%

phrases indicating conditional reasoning or causation

No Comments

No Known Activations