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

1:0.06

2:0.09

3:0.09

4:0.08

5:0.07

6:0.07

7:0.07

8:0.09

9:0.07

10:0.07

11:0.08

Negative Logits

 Athlet

-1.79

 Beth

-1.74

 workplaces

-1.62

 Meadows

-1.59

Bie

-1.58

 Deliver

-1.51

 Cases

-1.50

 Alto

-1.48

 concerts

-1.47

AGES

-1.47

POSITIVE LOGITS

ruption

1.85

erity

1.85

zai

1.77

ensor

1.73

condition

1.72

pri

1.72

protection

1.70

kered

1.66

ollah

1.65

imov

1.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.