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

1:0.08

2:0.08

3:0.08

4:0.07

5:0.07

6:0.08

7:0.08

8:0.08

9:0.09

10:0.09

11:0.07

Negative Logits

etheless

-2.09

received

-1.75

interstitial

-1.71

 incarcer

-1.70

encrypted

-1.70

replace

-1.68

ublished

-1.66

rint

-1.63

 recapt

-1.62

updated

-1.58

POSITIVE LOGITS

 Majesty

1.95

Magikarp

1.76

 Fathers

1.75

 objection

1.73

izons

1.71

POSE

1.61

Runner

1.59

 luckily

1.54

 hindsight

1.50

AppData

1.49

Activations Density 0.000%

No Known Activations

This feature has no known activations.