INDEX

Explanations

spatial relationships and directional terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.02

1:0.02

2:0.11

3:0.05

4:0.12

5:0.02

6:0.30

7:0.08

8:0.03

9:0.03

10:0.07

11:0.09

Negative Logits

 Architects

-1.31

Mex

-1.29

 inexper

-1.23

 respectively

-1.19

TBD

-1.19

 Sage

-1.18

wolves

-1.18

 Egyptians

-1.15

 Experiment

-1.15

 Ally

-1.14

POSITIVE LOGITS

rast

1.60

grounds

1.56

catentry

1.55

 stairs

1.53

atche

1.45

capacity

1.42

limits

1.37

quit

1.33

 tubes

1.33

aroo

1.31

Activations Density 0.003%

spatial relationships and directional terms

No Comments

No Known Activations