INDEX

Explanations

text related to online contests or submissions and fabricated stories or deceiving situations

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

ezvous

-0.65

 wonders

-0.63

thood

-0.63

blems

-0.63

 virginity

-0.63

ropolis

-0.61

urden

-0.61

 greeted

-0.60

 havoc

-0.60

utations

-0.60

POSITIVE LOGITS

 consortium

0.80

 Shap

0.76

 margin

0.72

 ausp

0.71

 Architects

0.70

 scrut

0.69

 Editors

0.68

 standpoint

0.67

 onwards

0.65

 collaborators

0.62

Activations Density 5.368%

text related to online contests or submissions and fabricated stories or deceiving situations

No Comments

No Known Activations