INDEX

Explanations

keywords related to military, technology, and society for future interpretation

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

rongh

-0.69

phabet

-0.67

mage

-0.67

ociate

-0.63

acan

-0.62

iverpool

-0.61

uli

-0.60

nants

-0.60

-0.59

iping

-0.58

POSITIVE LOGITS

 coincidence

0.80

 raining

0.76

ceivable

0.75

 folly

0.69

 impossible

0.66

to

0.65

 hypocritical

0.65

 uphill

0.61

 ironic

0.60

 enough

0.58

Activations Density 1.030%

keywords related to military, technology, and society for future interpretation

No Comments

No Known Activations