INDEX

Explanations

instances of content related to show or episode descriptions

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

1:0.03

2:0.04

3:0.10

4:0.05

5:0.04

6:0.26

7:0.04

8:0.05

9:0.06

10:0.08

11:0.15

Negative Logits

inav

-1.32

 sweat

-1.30

 caliphate

-1.16

bourg

-1.14

 Bradford

-1.11

 laure

-1.10

pload

-1.09

inals

-1.09

 culp

-1.07

IRD

-1.05

POSITIVE LOGITS

Merit

1.52

thora

1.45

 pione

1.26

endo

1.24

ItemImage

1.23

acly

1.22

theless

1.22

ヤ

1.19

 teasp

1.18

talk

1.18

Activations Density 0.000%

instances of content related to show or episode descriptions

No Comments

No Known Activations