INDEX

Explanations

instances of events and their corresponding dates or announcements

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

1:0.02

2:0.09

3:0.10

4:0.02

5:0.05

6:0.05

7:0.12

8:0.11

9:0.19

10:0.06

11:0.12

Negative Logits

ecake

-1.22

 Posts

-1.21

 insured

-1.14

onga

-1.13

aintain

-1.13

akia

-1.13

 vain

-1.13

 persisted

-1.12

umann

-1.10

uron

-1.08

POSITIVE LOGITS

 unveil

1.26

NAS

1.23

NYSE

1.23

 groundbreaking

1.21

 newcom

1.19

 revamped

1.16

版

1.16

 unveiling

1.16

 Billboard

1.14

Race

1.12

Activations Density 0.016%

instances of events and their corresponding dates or announcements

No Comments

No Known Activations