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

1:0.08

2:0.10

3:0.07

4:0.09

5:0.08

6:0.07

7:0.07

8:0.06

9:0.07

10:0.08

11:0.09

Negative Logits

oké

-3.33

Rating

-2.88

arial

-2.81

Final

-2.81

Transfer

-2.68

isSpecialOrderable

-2.66

Sleep

-2.64

ultimate

-2.59

 Offline

-2.58

 Goodbye

-2.49

POSITIVE LOGITS

FIA

2.77

2.46

Ake

2.46

agles

2.44

 Phill

2.43

 Michele

2.43

CFR

2.43

 sculpt

2.41

CAR

2.39

 ROCK

2.38

Activations Density 0.000%

No Known Activations

This feature has no known activations.