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

1:0.08

2:0.08

3:0.08

4:0.07

5:0.10

6:0.06

7:0.07

8:0.09

9:0.08

10:0.08

11:0.08

Negative Logits

ropolitan

-1.95

racuse

-1.75

Det

-1.71

apult

-1.66

veyard

-1.65

atorium

-1.64

thora

-1.63

ulty

-1.57

 Torch

-1.56

ynam

-1.54

POSITIVE LOGITS

 corrid

1.74

 monarch

1.70

 exempted

1.67

 pacif

1.67

 gubernatorial

1.66

 licens

1.58

 denim

1.58

assad

1.57

 promul

1.56

 millenn

1.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.