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

0:0.08

1:0.07

2:0.08

3:0.08

4:0.07

5:0.06

6:0.08

7:0.08

8:0.09

9:0.09

10:0.08

11:0.08

Negative Logits

 Surv

-2.77

 Bene

-2.51

bidden

-2.50

 unbeliev

-2.50

idon

-2.47

trap

-2.44

 Virtue

-2.44

-+-+

-2.43

uzzle

-2.42

 Catch

-2.38

POSITIVE LOGITS

Zac

2.88

assad

2.75

ulkan

2.57

apache

2.56

abad

2.55

 Kaplan

2.48

 Castro

2.46

 PowerPoint

2.41

jar

2.40

 curric

2.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.