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

2:0.09

3:0.06

4:0.10

5:0.09

6:0.07

7:0.08

8:0.08

9:0.08

10:0.07

11:0.08

Negative Logits

din

-1.76

dinand

-1.73

zilla

-1.69

ulence

-1.69

BuyableInstoreAndOnline

-1.65

nature

-1.62

 Vegeta

-1.58

inates

-1.57

ailability

-1.57

Bi

-1.56

POSITIVE LOGITS

 discouraged

1.89

 conscientious

1.73

earch

1.69

 regained

1.63

 Editorial

1.62

 resumed

1.59

 fortunes

1.58

 governed

1.58

 paralyzed

1.52

 unrem

1.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.