INDEX

Explanations

adjectives that express strong or negative qualities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

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

1:0.02

2:0.06

3:0.06

4:0.05

5:0.04

6:0.41

7:0.11

8:0.04

9:0.04

10:0.05

11:0.05

Negative Logits

DAY

-1.36

 Centauri

-1.23

 neighb

-1.21

ゴン

-1.21

 Dinosaur

-1.19

Za

-1.18

Pool

-1.10

 Jeanne

-1.10

 accuser

-1.10

Xia

-1.10

POSITIVE LOGITS

(>

1.57

arnaev

1.51

ersive

1.45

cientious

1.40

anto

1.39

rez

1.37

ileged

1.35

esta

1.35

 agric

1.34

tarian

1.31

Activations Density 0.012%

adjectives that express strong or negative qualities

No Comments

No Known Activations