INDEX

Explanations

the definite article "the"

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

1:0.09

2:0.07

3:0.07

4:0.09

5:0.08

6:0.10

7:0.08

8:0.07

9:0.08

10:0.08

11:0.07

Negative Logits

 Powers

-2.54

 Forge

-2.44

 Maker

-2.42

maker

-2.40

»

-2.36

Gim

-2.33

 Oliver

-2.33

 Reagan

-2.28

 Pistons

-2.24

 Norman

-2.23

POSITIVE LOGITS

trak

2.76

ugen

2.63

renheit

2.58

2.56

emale

2.54

 dorm

2.53

Wen

2.52

�

2.52

paio

2.51

 relegation

2.50

Activations Density 0.000%

the definite article "the"

No Comments

No Known Activations