INDEX

Explanations

phrases indicating progress or the anticipation of future actions

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

2:0.12

3:0.06

4:0.27

5:0.02

6:0.03

7:0.21

8:0.05

9:0.03

10:0.05

11:0.07

Negative Logits

edly

-1.51

SPONSORED

-1.49

Private

-1.46

ulhu

-1.38

Individual

-1.37

aneously

-1.35

rio

-1.32

pora

-1.31

 Administ

-1.30

thood

-1.29

POSITIVE LOGITS

 hurdle

1.78

abling

1.64

 underest

1.58

 equation

1.52

 underestimate

1.52

ynt

1.50

 regul

1.49

 adequ

1.45

 cutoff

1.43

 gears

1.41

Activations Density 0.001%

phrases indicating progress or the anticipation of future actions

No Comments

No Known Activations