INDEX

Explanations

No Explanations Found

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

0:0.09

1:0.06

2:0.07

3:0.08

4:0.07

5:0.09

6:0.08

7:0.08

8:0.08

9:0.07

10:0.08

11:0.08

Negative Logits

)</

-3.13

sel

-2.96

Loc

-2.72

 locating

-2.70

 colle

-2.58

@@

-2.44

stal

-2.38

ILL

-2.38

vill

-2.38

アル

-2.37

POSITIVE LOGITS

 Bahá

3.10

 circumcision

2.83

 Lenin

2.71

 preached

2.70

 anarchism

2.69

 aerobic

2.67

 Cohn

2.63

ascar

2.55

 offence

2.52

 circumcised

2.52

Activations Density 0.000%

No Known Activations

This feature has no known activations.