INDEX

Explanations

references to political elections and incumbency

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_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.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

0:0.05

1:0.14

2:0.07

3:0.09

4:0.08

5:0.02

6:0.16

7:0.11

8:0.04

9:0.05

10:0.04

11:0.10

Negative Logits

olkien

-2.62

 Scarlett

-2.55

raq

-2.54

Soc

-2.42

 Gaga

-2.38

 Canaver

-2.36

aphael

-2.36

Naz

-2.35

translation

-2.33

ryan

-2.32

POSITIVE LOGITS

 incumb

5.64

 incumbent

5.22

 challengers

3.97

 vacancies

3.52

��

3.17

 challenger

3.14

 districts

2.94

 outgoing

2.87

NBN

2.73

 vacancy

2.70

Activations Density 0.001%

references to political elections and incumbency

No Comments

No Known Activations