INDEX

Explanations

words related to cyber activities and discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ital

-0.76

 Lowry

-0.70

 Kenny

-0.70

 EntityItem

-0.69

 Lanc

-0.69

Klu

-0.69

 laure

-0.68

 Chiefs

-0.68

 Norris

-0.67

 Cherokee

-0.66

POSITIVE LOGITS

driven

1.33

based

1.33

mediated

1.26

centric

1.21

enabled

1.19

intensive

1.16

engineering

1.16

inspired

1.16

generated

1.14

sav

1.14

Activations Density 0.075%

words related to cyber activities and discussions

No Comments

No Known Activations