INDEX

Explanations

references to expediting or speeding up processes, especially related to reviews, permitting, and development in various contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases related to governmental infrastructure projects and expedited processes

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

words related to expedited processes and approvals within various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-att-kk

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.04

1:0.03

2:0.18

3:0.05

4:0.22

5:0.07

6:0.03

7:0.04

8:0.09

9:0.11

10:0.06

11:0.03

Negative Logits

 ILCS

-1.47

 likeness

-1.32

 resemblance

-1.32

龍�

-1.29

 裏�

-1.29

stocks

-1.25

Sync

-1.21

 upside

-1.20

 サーティワン

-1.18

bleacher

-1.17

POSITIVE LOGITS

hement

1.84

oided

1.33

 sailed

1.33

 Olympia

1.33

 Crusade

1.32

 dispatched

1.31

ctica

1.30

aneers

1.26

 undertaken

1.26

 disbanded

1.26

Activations Density 0.003%

references to expediting or speeding up processes, especially related to reviews, permitting, and development in various contexts

phrases related to governmental infrastructure projects and expedited processes

words related to expedited processes and approvals within various contexts

No Comments

No Known Activations