INDEX

Explanations

emphatic expressions or phrases indicating strong feelings or sentiments

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

New Auto-Interp

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

0:0.02

1:0.02

2:0.16

3:0.11

4:0.06

5:0.03

6:0.27

7:0.04

8:0.06

9:0.06

10:0.07

11:0.05

Negative Logits

 metropolitan

-1.29

 sandbox

-1.24

 Lancaster

-1.18

 remotely

-1.16

run

-1.16

 Peel

-1.15

 Hills

-1.15

 wilderness

-1.14

CFL

-1.13

 banking

-1.13

POSITIVE LOGITS

cffffcc

1.90

itely

1.59

andum

1.54

quote

1.45

govtrack

1.44

 Cosponsors

1.43

 VIDE

1.43

wcsstore

1.42

田

1.40

ername

1.38

Activations Density 0.057%

emphatic expressions or phrases indicating strong feelings or sentiments

No Comments

No Known Activations