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

1:0.05

2:0.09

3:0.09

4:0.06

5:0.08

6:0.09

7:0.09

8:0.08

9:0.09

10:0.08

11:0.08

Negative Logits

 berth

-1.87

wig

-1.84

 airing

-1.75

 Abrams

-1.65

Las

-1.62

 backstage

-1.60

 lens

-1.59

cis

-1.59

 actress

-1.57

MAC

-1.57

POSITIVE LOGITS

WARE

2.07

Reviewed

2.02

2.01

only

2.00

ernels

1.95

cycl

1.94

Topics

1.93

Adds

1.92

uggest

1.92

soever

1.88

Activations Density 0.000%

No Known Activations

This feature has no known activations.