INDEX

Explanations

No Explanations Found

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

0:0.08

1:0.06

2:0.09

3:0.09

4:0.06

5:0.08

6:0.08

7:0.08

8:0.08

9:0.08

10:0.09

11:0.08

Negative Logits

 Rebels

-1.44

 Firearms

-1.44

 Manga

-1.41

 Abrams

-1.41

 Corrections

-1.39

 Comics

-1.39

 Imaging

-1.37

 Transit

-1.37

apters

-1.36

isodes

-1.36

POSITIVE LOGITS

 repentance

1.92

 expectancy

1.83

ONSORED

1.69

 profitability

1.48

 motive

1.47

 greed

1.42

 gradual

1.42

fam

1.40

 remorse

1.39

 inaction

1.37

Activations Density 0.000%

No Known Activations

This feature has no known activations.