INDEX

Explanations

URLs and web links, especially from Twitter

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

1:0.03

2:0.10

3:0.12

4:0.27

5:0.03

6:0.02

7:0.17

8:0.02

9:0.03

10:0.09

11:0.06

Negative Logits

phabet

-1.44

 proofs

-1.38

 preparations

-1.35

 interstellar

-1.34

 worthy

-1.32

sonian

-1.31

 upgrades

-1.30

 trophies

-1.29

 pathology

-1.28

CRIPTION

-1.28

POSITIVE LOGITS

ifier

1.59

say

1.46

nih

1.44

users

1.40

comment

1.39

 sacked

1.39

df

1.39

eq

1.39

yn

1.37

hash

1.36

Activations Density 0.003%

URLs and web links, especially from Twitter

No Comments

No Known Activations