INDEX

Explanations

terms related to transactions, interactions, or exchanges involving financial or physical goods or services

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Airl

-0.69

MW

-0.65

 Revival

-0.63

ondo

-0.63

 Rite

-0.61

roman

-0.61

 Pike

-0.60

 situational

-0.60

 Cooking

-0.60

inventoryQuantity

-0.60

POSITIVE LOGITS

hip

1.27

hips

1.09

 beware

1.06

paces

0.94

who

0.92

ettings

0.91

 whom

0.90

 alike

0.85

 thereof

0.83

who

0.80

Activations Density 0.124%

terms related to transactions, interactions, or exchanges involving financial or physical goods or services

No Comments

No Known Activations