INDEX

Explanations

phrases related to confrontation and personal accountability in discussions

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_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.3.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

0:0.01

1:0.03

2:0.08

3:0.06

4:0.02

5:0.03

6:0.06

7:0.09

8:0.25

9:0.08

10:0.11

11:0.12

Negative Logits

iggins

-1.18

elo

-1.16

�

-1.09

prise

-1.09

�

-1.08

ipel

-1.04

adelphia

-1.01

eger

-1.00

ás

-1.00

alg

-0.99

POSITIVE LOGITS

 hadn

1.14

');

1.12

')

1.09

 clicked

1.02

 discriminated

1.01

'),

1.00

 existed

1.00

 didnt

0.99

 lied

0.96

chwitz

0.95

Activations Density 0.003%

phrases related to confrontation and personal accountability in discussions

No Comments

No Known Activations