INDEX

Explanations

references to geopolitical entities or divisions, particularly related to regions or cities and their socio-economic 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

0:0.07

1:0.01

2:0.05

3:0.52

4:0.02

5:0.09

6:0.02

7:0.03

8:0.05

9:0.01

10:0.07

11:0.01

Negative Logits

 Lifetime

-1.88

STD

-1.87

 professionalism

-1.83

Professional

-1.81

 referees

-1.75

 warranties

-1.75

Listener

-1.74

Statement

-1.73

 assistants

-1.72

 invitations

-1.70

POSITIVE LOGITS

 countryside

3.58

 southeast

3.58

 northwest

3.50

 suburbs

3.50

 southeastern

3.49

 northeast

3.48

 northwestern

3.45

 northern

3.45

 suburb

3.45

region

3.39

Activations Density 2.225%

references to geopolitical entities or divisions, particularly related to regions or cities and their socio-economic contexts

No Comments

No Known Activations