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

0:0.09

1:0.07

2:0.09

3:0.08

4:0.07

5:0.08

6:0.08

7:0.09

8:0.08

9:0.06

10:0.08

11:0.08

Negative Logits

Sov

-3.12

 compr

-3.12

odynam

-2.92

doctor

-2.86

DNA

-2.86

MRI

-2.85

ieft

-2.81

immune

-2.80

 replacements

-2.77

engine

-2.76

POSITIVE LOGITS

 neighbourhoods

3.21

 annex

3.20

 Finch

3.19

Mul

3.10

 Sophie

3.09

 Edinburgh

3.06

 Idle

3.04

 Hastings

2.92

 Brook

2.86

 Goodman

2.86

Activations Density 0.000%

No Known Activations

This feature has no known activations.