INDEX

Explanations

percentages and statistics that involve people

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.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

0:0.07

1:0.04

2:0.09

3:0.13

4:0.04

5:0.14

6:0.04

7:0.04

8:0.11

9:0.08

10:0.12

11:0.05

Negative Logits

eus

-1.35

zik

-1.21

 PAGE

-1.14

 translator

-1.08

��

-1.08

eger

-1.07

 envoy

-1.06

 grate

-1.05

ゴン

-1.03

 Ramsay

-1.02

POSITIVE LOGITS

 polled

1.59

total

1.33

 surveyed

1.26

 households

1.24

depending

1.20

 Including

1.13

excluding

1.10

eligible

1.08

 Adults

1.06

insured

1.06

Activations Density 0.109%

percentages and statistics that involve people

No Comments

No Known Activations