INDEX

Explanations

negations or phrases indicating impossibility or challenges

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

0:0.03

1:0.03

2:0.04

3:0.06

4:0.01

5:0.04

6:0.08

7:0.05

8:0.10

9:0.24

10:0.12

11:0.15

Negative Logits

 antioxid

-1.16

 Tomorrow

-1.13

��

-1.11

……………………

-1.07

agan

-1.06

uthor

-0.98

%%%%

-0.94

 Brands

-0.94

FORMATION

-0.93

ramid

-0.91

POSITIVE LOGITS

 refreshing

1.03

ounters

1.03

ners

1.02

 Deng

1.01

hess

0.97

ressed

0.95

ted

0.93

 resur

0.92

apy

0.91

requently

0.90

Activations Density 0.067%

negations or phrases indicating impossibility or challenges

No Comments

No Known Activations