INDEX

Explanations

phrases indicating coercion and manipulation in various contexts

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_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

0:0.05

1:0.02

2:0.07

3:0.13

4:0.02

5:0.07

6:0.01

7:0.06

8:0.02

9:0.02

10:0.46

11:0.01

Negative Logits

 resemblance

-2.13

oire

-2.11

ibaba

-2.05

 millenn

-2.05

 similarities

-1.97

aez

-1.95

ーティ

-1.91

iaries

-1.90

 illust

-1.88

nesota

-1.88

POSITIVE LOGITS

 acquies

2.71

 compromise

2.71

 compromises

2.67

 concessions

2.61

 sacrifices

2.33

 obedience

2.22

 outright

2.20

 obey

2.16

 behave

2.10

 resisted

2.09

Activations Density 0.289%

phrases indicating coercion and manipulation in various contexts

No Comments

No Known Activations