INDEX

Explanations

statements relating to quantity or measurement

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

Head Attr Weights

0:0.08

1:0.08

2:0.07

3:0.08

4:0.07

5:0.08

6:0.08

7:0.09

8:0.08

9:0.07

10:0.09

11:0.07

Negative Logits

Pac

-2.88

azard

-2.88

outheast

-2.83

SCP

-2.82

illus

-2.81

emphasis

-2.75

Pac

-2.73

��

-2.66

aquin

-2.61

Armor

-2.59

POSITIVE LOGITS

 Sochi

3.33

 petrol

2.85

 simul

2.83

 tomorrow

2.82

 Seym

2.75

dyn

2.69

Spa

2.65

 champagne

2.61

 Wembley

2.56

 estab

2.56

Activations Density 0.000%

statements relating to quantity or measurement

No Known Activations