INDEX

Explanations

instances of abrupt changes or moments of realization

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_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.3.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

1:0.05

2:0.10

3:0.04

4:0.02

5:0.05

6:0.06

7:0.14

8:0.08

9:0.11

10:0.05

11:0.24

Negative Logits

 Revolutionary

-1.22

 endorsements

-1.19

 Australian

-1.17

 hence

-1.14

 transmitting

-1.12

 Ulster

-1.12

 sorely

-1.11

 Indies

-1.11

 Declaration

-1.10

 equally

-1.09

POSITIVE LOGITS

iphany

1.53

yss

1.47

 earthqu

1.46

amaz

1.41

ADRA

1.40

raged

1.39

uclear

1.37

ycle

1.35

uder

1.35

arrass

1.35

Activations Density 0.003%

instances of abrupt changes or moments of realization

No Comments

No Known Activations