INDEX

Explanations

mentions of specific locations, particularly focusing on "New York."

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

ONSORED

-0.91

ï¸ı

-0.81

 actionGroup

-0.71

pless

-0.71

 sqor

-0.69

xual

-0.69

cius

-0.68

uca

-0.67

ional

-0.66

ILA

-0.66

POSITIVE LOGITS

 York

1.55

 Zealand

1.48

 Orleans

1.40

 Hampshire

1.28

 Jersey

1.21

 Yorker

1.17

 Yorkers

1.14

York

1.13

 Testament

1.10

 Brunswick

1.08

Activations Density 0.463%

mentions of specific locations, particularly focusing on "New York."

No Comments

No Known Activations