INDEX

Explanations

phrases related to documentation or procedural content

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

0:0.02

1:0.04

2:0.07

3:0.11

4:0.02

5:0.05

6:0.09

7:0.29

8:0.06

9:0.05

10:0.05

11:0.09

Negative Logits

vard

-1.34

enegger

-1.29

 Johannes

-1.19

Jed

-1.18

MPH

-1.14

 Bots

-1.12

 coli

-1.10

ahime

-1.05

ester

-1.03

urdue

-1.03

POSITIVE LOGITS

��

1.42

版

1.32

 luggage

1.24

ufact

1.22

aque

1.12

ナ

1.12

ヴァ

1.11

Item

1.08

 ensu

1.08

Compan

1.08

Activations Density 0.027%

phrases related to documentation or procedural content

No Comments

No Known Activations