INDEX

Explanations

references to sports and competition

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_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

0:0.08

1:0.03

2:0.06

3:0.03

4:0.05

5:0.05

6:0.25

7:0.05

8:0.08

9:0.22

10:0.02

11:0.04

Negative Logits

Lenin

-4.07

 Uran

-4.03

�

-3.70

Georg

-3.64

 Lenin

-3.51

 toast

-3.44

Dig

-3.42

 fortress

-3.42

 phil

-3.38

 horizont

-3.35

POSITIVE LOGITS

 Riley

11.90

 Raleigh

5.35

 Rodney

5.15

iley

5.13

 Irvine

4.83

 Rice

4.82

 Reeves

4.66

RI

4.53

 Payton

4.26

Rae

4.18

Activations Density 0.003%

references to sports and competition

No Comments

No Known Activations