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

1:0.07

2:0.09

3:0.08

4:0.08

5:0.09

6:0.08

7:0.07

8:0.09

9:0.07

10:0.07

11:0.07

Negative Logits

 Abrams

-2.02

 Univ

-1.95

 Southwest

-1.88

 Jace

-1.84

 Cheng

-1.76

Mex

-1.74

 istg

-1.73

 Brus

-1.73

Cul

-1.73

Ago

-1.73

POSITIVE LOGITS

Untitled

2.07

rain

1.87

plet

1.84

visory

1.83

gency

1.82

obin

1.78

spection

1.75

UTE

1.75

ker

1.71

blocking

1.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.