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

1:0.08

2:0.08

3:0.07

4:0.08

5:0.07

6:0.07

7:0.09

8:0.08

9:0.09

10:0.08

11:0.08

Negative Logits

Rogue

-3.20

Limit

-2.96

quished

-2.95

elsh

-2.78

DoS

-2.76

 Slayer

-2.76

atche

-2.75

rikes

-2.70

rast

-2.65

ifted

-2.62

POSITIVE LOGITS

 Femin

2.66

 documenting

2.57

 feminist

2.46

 corrobor

2.44

 Patri

2.43

models

2.42

 Feminist

2.41

 Easter

2.40

 presentation

2.37

 bore

2.36

Activations Density 0.000%

No Known Activations

This feature has no known activations.