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

4:0.09

5:0.08

6:0.08

7:0.08

8:0.06

9:0.08

10:0.09

11:0.08

Negative Logits

dark

-2.68

Todd

-2.66

ASHINGTON

-2.65

FRI

-2.56

~~~~~~~~

-2.44

________________________________

-2.40

>(

-2.40

�

-2.39

follow

-2.38

Princ

-2.37

POSITIVE LOGITS

AMA

2.91

 impeachment

2.79

 Napoleon

2.73

pless

2.65

ieri

2.59

Interstitial

2.58

Aut

2.57

Amp

2.51

Battery

2.46

ibaba

2.43

Activations Density 0.000%

No Known Activations

This feature has no known activations.