INDEX

Explanations

expressions of appreciation and gratitude

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ificantly

-0.63

 contradicted

-0.62

 pred

-0.59

SPONSORED

-0.59

conserv

-0.59

 delinqu

-0.59

igmat

-0.59

Compare

-0.58

uffer

-0.57

atton

-0.56

POSITIVE LOGITS

 congr

1.05

!!!

1.00

 welcome

0.98

 congratulations

0.98

sir

0.97

!!!!!

0.96

 thank

0.95

:)

0.94

!!!!

0.93

Activations Density 2.555%

expressions of appreciation and gratitude

No Comments

No Known Activations