INDEX

Explanations

specific mentions of military units, divisions, or numerical references

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 Monsters

-0.66

Statistics

-0.65

bos

-0.63

uploads

-0.62

mbuds

-0.61

Downloadha

-0.60

info

-0.60

bda

-0.60

values

-0.59

Settings

-0.59

POSITIVE LOGITS

ousand

1.15

 century

1.07

 inning

1.07

irty

1.06

 percentile

1.01

century

0.97

 Amendment

0.96

 installment

0.91

ousands

0.91

 Century

0.90

Activations Density 2.087%

specific mentions of military units, divisions, or numerical references

No Comments

No Known Activations