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

4:0.09

5:0.08

6:0.08

7:0.09

8:0.08

9:0.08

10:0.07

11:0.08

Negative Logits

 Osiris

-1.99

thood

-1.98

ufact

-1.79

 Owners

-1.75

 Includes

-1.68

atin

-1.61

 Proper

-1.58

selves

-1.57

 Catholics

-1.56

 Territories

-1.55

POSITIVE LOGITS

 bullish

1.71

USD

1.70

 disappoint

1.67

 catapult

1.57

 brainstorm

1.53

OND

1.50

 democrat

1.50

 accelerator

1.48

HK

1.47

 whiff

1.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.