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

1:0.09

2:0.08

3:0.08

4:0.08

5:0.08

6:0.09

7:0.08

8:0.07

9:0.07

10:0.07

11:0.08

Negative Logits

 museums

-2.99

 nonex

-2.85

 mosquito

-2.77

 mayors

-2.75

 planes

-2.71

‎

-2.69

 architect

-2.67

 bount

-2.67

 livest

-2.62

 conven

-2.61

POSITIVE LOGITS

JB

3.22

 Lesbian

2.98

 Barker

2.90

ORY

2.86

Bell

2.85

 Punk

2.83

 Hayes

2.80

 Sisters

2.76

 Shaman

2.75

HW

2.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.