INDEX

Explanations

short phrases or sentences that allude to negative aspects or events

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Cash

-0.69

 Pharm

-0.63

Tes

-0.62

Weather

-0.60

Medical

-0.60

Kar

-0.58

eli

-0.57

Trust

-0.57

Kill

-0.56

 Virgin

-0.56

POSITIVE LOGITS

 designation

0.76

 signifies

0.76

 entails

0.74

 altogether

0.73

 anyway

0.73

 refers

0.72

 equals

0.71

 implies

0.71

0.70

 exists

0.69

Activations Density 5.414%

short phrases or sentences that allude to negative aspects or events

No Comments

No Known Activations