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

Head Attr Weights

0:0.06

1:0.09

2:0.08

3:0.07

4:0.08

5:0.05

6:0.11

7:0.09

8:0.09

9:0.08

10:0.08

11:0.07

Negative Logits

ileaks

-1.68

 Gandhi

-1.66

 Tomato

-1.62

ilib

-1.59

 Harden

-1.58

ilian

-1.55

acan

-1.54

 architectures

-1.54

leaders

-1.54

 Hipp

-1.53

POSITIVE LOGITS

 permanently

1.69

channelAvailability

1.69

 confined

1.65

ヴ

1.59

 indefinitely

1.57

 longer

1.56

 secondly

1.54

 newcomers

1.53

iguous

1.45

 forcefully

1.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.