INDEX

Explanations

technical and legal terms related to intellectual property, pharmacy, and science

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

swick

-0.71

UAL

-0.67

 Canary

-0.67

nings

-0.66

Interstitial

-0.65

 Olson

-0.64

ITED

-0.64

vier

-0.63

 Wend

-0.63

FORE

-0.62

POSITIVE LOGITS

odynamic

1.36

odynamics

1.33

ologically

1.31

ysics

1.27

odynam

1.25

osph

1.25

ographical

1.24

ocry

1.22

ological

1.22

oton

1.20

Activations Density 4.628%

technical and legal terms related to intellectual property, pharmacy, and science

No Comments

No Known Activations