INDEX

Explanations

negative statements or sentiments

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

 interval

-0.75

 Compliance

-0.68

ND

-0.67

 exit

-0.66

KP

-0.66

 conduc

-0.65

 Rouhani

-0.64

 Govern

-0.64

FB

-0.64

UD

-0.64

POSITIVE LOGITS

shaped

1.35

filled

1.22

faced

1.20

headed

1.18

loving

1.16

covered

1.16

like

1.14

clad

1.13

themed

1.11

eyed

1.11

Activations Density 0.094%

negative statements or sentiments

No Comments

No Known Activations