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

2:0.08

3:0.09

4:0.08

5:0.08

6:0.06

7:0.09

8:0.10

9:0.08

10:0.09

11:0.07

Negative Logits

inav

-2.13

ovan

-1.85

items

-1.78

Jews

-1.69

oslov

-1.66

atin

-1.66

ingo

-1.66

ukong

-1.65

requ

-1.64

agen

-1.64

POSITIVE LOGITS

 whistlebl

1.89

 dips

1.79

emouth

1.77

 premie

1.76

 tumble

1.66

 lows

1.66

 airs

1.65

 helm

1.59

 recl

1.59

 examines

1.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.