INDEX

Explanations

increases or changes in position or direction

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

Interstitial

-0.80

gom

-0.72

ORED

-0.70

gdala

-0.70

gil

-0.67

CRIP

-0.66

amina

-0.66

 Seym

-0.65

ortium

-0.64

çĦ

-0.63

POSITIVE LOGITS

 gears

1.13

 shift

0.92

 sands

0.92

 shifts

0.83

 shifting

0.83

 away

0.82

otom

0.82

olitics

0.80

over

0.79

 toward

0.78

Activations Density 0.029%

increases or changes in position or direction

No Comments

No Known Activations