INDEX

Explanations

words related to cybersecurity attacks and different kinds of crimes

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

 Sabha

-0.76

BIL

-0.71

OTAL

-0.70

 rall

-0.69

entary

-0.68

 tert

-0.67

CHO

-0.64

 propri

-0.63

FAULT

-0.62

uez

-0.62

POSITIVE LOGITS

haven

0.90

hemy

0.88

headed

0.83

feeding

0.82

eworthy

0.82

blooded

0.78

pedia

0.78

fast

0.75

 trout

0.74

mouth

0.72

Activations Density 0.053%

words related to cybersecurity attacks and different kinds of crimes

No Comments

No Known Activations