INDEX

Explanations

instances of actions or outcomes related to avoiding, minimizing, or handling risks and consequences

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

0:0.02

1:0.01

2:0.05

3:0.08

4:0.10

5:0.03

6:0.40

7:0.06

8:0.05

9:0.04

10:0.05

11:0.05

Negative Logits

ALLY

-1.45

ォ

-1.41

田

-1.39

yip

-1.29

 Username

-1.27

stals

-1.25

loo

-1.24

 somew

-1.23

▬

-1.23

utra

-1.21

POSITIVE LOGITS

 altogether

1.68

 vomiting

1.39

effects

1.38

 coerc

1.35

oxicity

1.35

 starvation

1.34

akable

1.32

 effectiveness

1.31

 restraints

1.31

ogun

1.30

Activations Density 0.010%

instances of actions or outcomes related to avoiding, minimizing, or handling risks and consequences

No Comments

No Known Activations