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

3:0.09

4:0.07

5:0.08

6:0.07

7:0.08

8:0.09

9:0.08

10:0.09

11:0.07

Negative Logits

 wires

-3.29

 worms

-2.96

 wire

-2.86

othal

-2.85

 tether

-2.84

ileaks

-2.81

Dyn

-2.73

 cables

-2.67

 hacker

-2.59

 Rothschild

-2.58

POSITIVE LOGITS

FTA

2.97

 Spur

2.65

angered

2.64

unda

2.62

amaz

2.61

cussion

2.59

 Mour

2.58

 Kamp

2.55

iberal

2.53

afa

2.51

Activations Density 0.000%

No Known Activations

This feature has no known activations.