INDEX

Explanations

words related to protection, defense, and security

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

hall

-0.70

jet

-0.67

LINE

-0.64

LIN

-0.63

lore

-0.63

hler

-0.61

lins

-0.61

zos

-0.59

hyp

-0.57

 Minutes

-0.57

POSITIVE LOGITS

 against

1.21

against

1.10

iveness

1.09

ively

1.03

 Against

0.98

ously

0.97

atively

0.96

folios

0.92

orate

0.91

ailability

0.88

Activations Density 0.796%

words related to protection, defense, and security

No Comments

No Known Activations