INDEX

Explanations

phrases related to restaurants and food chains

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

potion

-0.74

CHAT

-0.74

tein

-0.70

Timer

-0.67

othy

-0.65

otos

-0.65

Accessory

-0.61

tun

-0.60

 beard

-0.60

Widget

-0.60

POSITIVE LOGITS

hips

1.07

chool

1.03

 specializing

0.97

 vying

0.96

 alike

0.94

 catering

0.94

 worldwide

0.91

paces

0.90

hare

0.89

 employing

0.89

Activations Density 4.791%

phrases related to restaurants and food chains

No Comments

No Known Activations