INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

3:0.07

4:0.10

5:0.07

6:0.08

7:0.08

8:0.07

9:0.08

10:0.08

11:0.09

Negative Logits

 Appearance

-1.71

 Incarnation

-1.65

 Alternate

-1.61

 Severus

-1.61

 Cups

-1.59

 Blades

-1.55

Rak

-1.55

 Background

-1.53

 Seoul

-1.52

cade

-1.52

POSITIVE LOGITS

覚醒

2.15

icans

2.10

APD

1.89

ïve

1.70

 doub

1.59

ッド

1.57

ensable

1.55

rave

1.53

 journalism

1.52

�

1.52

Activations Density 0.000%

No Known Activations

This feature has no known activations.