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

2:0.08

3:0.09

4:0.08

5:0.08

6:0.08

7:0.08

8:0.09

9:0.08

10:0.09

11:0.06

Negative Logits

cyl

-2.10

ymes

-1.85

rase

-1.82

 mater

-1.74

nan

-1.72

Sac

-1.65

 alike

-1.65

 infer

-1.63

 perpendicular

-1.61

 Xeon

-1.56

POSITIVE LOGITS

Street

1.62

irement

1.57

Folder

1.57

GROUP

1.56

IRED

1.54

Dem

1.53

IRD

1.51

 Gaga

1.48

Angelo

1.48

 unsuccessful

1.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.