INDEX

Explanations

occurrences of the word "to" in varying contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

words indicating actions or infinitive forms of verbs

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-att-kk

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

0:0.09

1:0.03

2:0.06

3:0.07

4:0.14

5:0.08

6:0.06

7:0.02

8:0.20

9:0.08

10:0.09

11:0.03

Negative Logits

 gulf

-1.31

ulty

-1.27

 mysteries

-1.15

 benef

-1.15

Must

-1.11

Gap

-1.11

ominated

-1.11

iated

-1.09

ior

-1.08

 surrounds

-1.08

POSITIVE LOGITS

ドラ

1.49

�

1.48

 seriously

1.43

*/(

1.41

 offline

1.40

 lightly

1.30

 photograph

1.29

essage

1.25

 Hispan

1.25

 Seriously

1.22

Activations Density 0.011%

occurrences of the word "to" in varying contexts

words indicating actions or infinitive forms of verbs

No Comments

No Known Activations