INDEX

Explanations

phrases related to social issues and conflicts

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

reon

-0.70

 largeDownload

-0.68

chwitz

-0.67

arching

-0.65

 Rousse

-0.63

ilee

-0.62

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.62

ophen

-0.61

 Indra

-0.60

VERTISEMENT

-0.59

POSITIVE LOGITS

pox

1.29

(<

0.98

 increments

0.92

 consolation

0.90

 tweaks

0.88

atur

0.87

 insignificant

0.86

fry

0.85

 handful

0.84

 incremental

0.83

Activations Density 0.646%

phrases related to social issues and conflicts

No Comments

No Known Activations