INDEX

Explanations

references to individuals or groups in various contexts

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

references and phrases that indicate attribution or direction

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-att-kk

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

Head Attr Weights

0:0.01

1:0.04

2:0.18

3:0.14

4:0.01

5:0.04

6:0.05

7:0.07

8:0.05

9:0.18

10:0.07

11:0.10

Negative Logits

 vying

-1.05

 resil

-0.99

 thri

-0.97

edom

-0.97

uffer

-0.92

 certs

-0.92

 Samoa

-0.89

think

-0.88

 Canad

-0.88

 thrive

-0.88

POSITIVE LOGITS

ール

1.12

ェ

1.11

ム

1.07

 Piercing

1.07

 empt

1.05

Suggest

1.03

��

1.03

ーテ

1.02

 insults

1.01

ジ

0.99

Activations Density 0.009%

references to individuals or groups in various contexts

references and phrases that indicate attribution or direction

No Known Activations

references to individuals or groups in various contexts

references and phrases that indicate attribution or direction

No Known Activations