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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.09

2:0.07

3:0.08

4:0.09

5:0.07

6:0.08

7:0.08

8:0.08

9:0.08

10:0.08

11:0.06

Negative Logits

anka

-3.22

DragonMagazine

-3.00

 Punjab

-2.88

 Ethiop

-2.85

 Indians

-2.82

chuk

-2.82

 Taliban

-2.71

entin

-2.68

 turb

-2.64

odan

-2.64

POSITIVE LOGITS

 payoff

2.85

 Pinball

2.77

 Vapor

2.69

 Grow

2.67

 crochet

2.58

 Customer

2.56

 patiently

2.48

 Pound

2.46

 Satisf

2.45

cog

2.44

Activations Density 0.000%

No Known Activations

This feature has no known activations.