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

0:0.10

1:0.07

2:0.07

3:0.09

4:0.07

5:0.08

6:0.08

7:0.07

8:0.09

9:0.08

10:0.09

11:0.07

Negative Logits

 senseless

-1.69

inges

-1.68

 brink

-1.67

 outburst

-1.67

]);

-1.65

 precip

-1.63

 meltdown

-1.59

 chorus

-1.57

 unprepared

-1.55

 backward

-1.53

POSITIVE LOGITS

Register

1.88

Beta

1.72

Reviewed

1.70

assis

1.68

ARC

1.68

Tickets

1.67

minimum

1.65

Assistant

1.63

WARE

1.61

YA

1.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.