INDEX

Explanations

specific names of teams and players in sports contexts

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

0:0.04

1:0.19

2:0.04

3:0.06

4:0.04

5:0.04

6:0.15

7:0.05

8:0.03

9:0.16

10:0.04

11:0.12

Negative Logits

 Carpenter

-3.45

 Fleming

-3.42

Cole

-3.10

 Parsons

-3.10

obs

-3.07

 Myers

-2.99

cientious

-2.89

━

-2.86

 Loving

-2.80

 Benson

-2.79

POSITIVE LOGITS

 versus

4.09

ieth

3.24

utral

2.79

 defeat

2.66

 defeating

2.64

Whereas

2.64

uyomi

2.60

 against

2.59

 suscept

2.56

wave

2.55

Activations Density 0.001%

specific names of teams and players in sports contexts

No Comments

No Known Activations