INDEX

Explanations

phrases related to physical actions and descriptions of scenes

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

 specialization

-0.79

 promotions

-0.76

 outsourcing

-0.70

 promotion

-0.69

 budgets

-0.68

MQ

-0.67

Rum

-0.67

 collaborations

-0.66

 demographics

-0.65

 economies

-0.63

POSITIVE LOGITS

stretched

1.17

 startled

1.07

 sideways

1.00

 perpendicular

0.99

 upright

0.99

 perpend

0.98

 protr

0.96

 crou

0.94

 shoulder

0.94

 crotch

0.94

Activations Density 3.517%

phrases related to physical actions and descriptions of scenes

No Comments

No Known Activations