INDEX

Explanations

words and phrases indicating causation or relationships among entities and concepts

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

1:0.04

2:0.07

3:0.05

4:0.02

5:0.04

6:0.04

7:0.05

8:0.09

9:0.41

10:0.07

11:0.06

Negative Logits

swer

-1.45

ify

-1.42

ITED

-1.36

ENG

-1.35

agle

-1.34

otiation

-1.34

ainer

-1.30

女

-1.28

oyer

-1.25

again

-1.24

POSITIVE LOGITS

 Chilean

1.34

enza

1.33

 mitigating

1.32

 Deaths

1.28

ieved

1.23

govtrack

1.22

 Manson

1.21

Gou

1.17

 Okin

1.15

ーティ

1.15

Activations Density 0.008%

words and phrases indicating causation or relationships among entities and concepts

No Comments

No Known Activations