INDEX

Explanations

references to a specific subject matter or context

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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.01

2:0.14

3:0.10

4:0.15

5:0.03

6:0.05

7:0.21

8:0.05

9:0.04

10:0.09

11:0.06

Negative Logits

inous

-1.51

 darts

-1.34

 hemor

-1.27

 dinosaurs

-1.23

 Jehovah

-1.21

 surging

-1.19

 axes

-1.19

 trolls

-1.18

 ridden

-1.18

 bubbles

-1.18

POSITIVE LOGITS

ermanent

1.45

transfer

1.39

1.34

minus

1.32

1.30

NP

1.25

ign

1.25

Required

1.25

=-=-

1.24

eff

1.24

Activations Density 0.001%

references to a specific subject matter or context

No Comments

No Known Activations