INDEX

Explanations

names of people involved in a conflict or altercation

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

category

-0.76

 endorsements

-0.70

actionDate

-0.70

 Individuals

-0.65

ishable

-0.65

interstitial

-0.65

 Pione

-0.64

Prem

-0.63

Effective

-0.63

natural

-0.62

POSITIVE LOGITS

 hesitated

1.24

 panicked

1.19

 awoke

1.19

 replied

1.16

 proceeded

1.16

 grabbed

1.13

 laughed

1.13

 responded

1.12

 yelled

1.09

 screamed

1.08

Activations Density 0.578%

names of people involved in a conflict or altercation

No Comments

No Known Activations