INDEX

Explanations

sport-related content, particularly baseball and football

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

0:0.06

1:0.07

2:0.05

3:0.08

4:0.03

5:0.08

6:0.04

7:0.07

8:0.04

9:0.23

10:0.14

11:0.07

Negative Logits

 compan

-1.35

 Shutterstock

-1.21

 reproduction

-1.17

bara

-1.12

 google

-1.11

 paras

-1.11

ie

-1.10

 suffice

-1.10

 Feminist

-1.10

 summar

-1.08

POSITIVE LOGITS

 Coliseum

1.81

ukong

1.71

 Playoffs

1.68

 preseason

1.61

venge

1.61

 Playoff

1.55

士

1.53

 playoff

1.52

 Heisman

1.45

 playoffs

1.45

Activations Density 0.232%

sport-related content, particularly baseball and football

No Comments

No Known Activations