INDEX

Explanations

references to time and clocks

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_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

0:0.35

1:0.02

2:0.18

3:0.09

4:0.03

5:0.06

6:0.03

7:0.03

8:0.02

9:0.04

10:0.05

11:0.02

Negative Logits

packs

-2.60

afort

-2.57

crop

-2.35

packing

-2.35

 waivers

-2.33

 Scholarship

-2.26

Medic

-2.24

 hospitality

-2.23

Rus

-2.21

ricanes

-2.21

POSITIVE LOGITS

 clock

4.62

 Clock

4.53

 countdown

4.37

 ticking

4.30

 clocks

4.30

 timer

4.18

clock

3.99

Clock

3.97

 alarm

3.85

 alarms

3.58

Activations Density 0.038%

references to time and clocks

No Comments

No Known Activations