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

0:0.07

1:0.07

2:0.09

3:0.08

4:0.09

5:0.08

6:0.08

7:0.07

8:0.08

9:0.07

10:0.08

11:0.08

Negative Logits

 Translation

-3.63

Ng

-2.81

translation

-2.77

 Verse

-2.76

tera

-2.63

ño

-2.57

 Literature

-2.57

quez

-2.53

Nun

-2.53

Buk

-2.50

POSITIVE LOGITS

 Costco

2.77

 dining

2.58

 behavi

2.52

 socially

2.48

 athlet

2.46

 departing

2.46

 corros

2.43

 college

2.41

ackle

2.33

 sever

2.31

Activations Density 0.000%

No Known Activations

This feature has no known activations.