INDEX

Explanations

phrases that involve the word "in" followed by varying context words, indicating a focus on the concept of inclusion or involvement in different scenarios

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_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

0:0.01

1:0.03

2:0.07

3:0.08

4:0.11

5:0.03

6:0.04

7:0.35

8:0.03

9:0.04

10:0.08

11:0.09

Negative Logits

ipment

-1.70

DragonMagazine

-1.64

 pend

-1.53

ubes

-1.50

ubs

-1.50

ebook

-1.45

assadors

-1.45

pite

-1.45

apest

-1.44

reetings

-1.41

POSITIVE LOGITS

 criminality

1.72

 terrorism

1.55

 wrongdoing

1.49

 causal

1.44

 existential

1.44

 paranormal

1.43

 identity

1.41

 criminal

1.39

 supernatural

1.37

 conspiracy

1.35

Activations Density 0.001%

phrases that involve the word "in" followed by varying context words, indicating a focus on the concept of inclusion or involvement in different scenarios

No Comments

No Known Activations