INDEX

Explanations

references to drugs and drug-related terminology in text

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

 convergence

-0.71

yrinth

-0.69

 Schro

-0.66

Tro

-0.63

Tun

-0.61

 Missile

-0.60

Weather

-0.59

ãĤ·ãĥ£

-0.58

Lob

-0.58

idth

-0.57

POSITIVE LOGITS

 willingly

1.05

 unlawfully

1.04

 illegally

1.03

 indoors

1.03

 oneself

1.01

 anytime

1.01

 abroad

1.01

 anymore

0.99

 professionally

0.98

selves

0.97

Activations Density 0.320%

references to drugs and drug-related terminology in text

No Comments

No Known Activations