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

0:0.08

1:0.08

2:0.07

3:0.08

4:0.09

5:0.08

6:0.08

7:0.09

8:0.09

9:0.07

10:0.07

11:0.07

Negative Logits

hews

-2.77

PDF

-2.75

 Americ

-2.68

Sah

-2.64

artifacts

-2.61

oops

-2.61

IPS

-2.59

oos

-2.58

Either

-2.52

=-=-

-2.52

POSITIVE LOGITS

 Rider

2.97

Cov

2.76

 Engel

2.74

 Stevenson

2.72

 subscription

2.68

2.67

 Clinical

2.67

netflix

2.64

 Trials

2.64

 inquest

2.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.