INDEX

Explanations

terms related to fire incidents

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

xus

-0.81

verty

-0.69

eter

-0.68

los

-0.68

DonaldTrump

-0.67

edy

-0.67

pai

-0.65

istan

-0.64

************

-0.63

kees

-0.63

POSITIVE LOGITS

flies

1.03

balls

1.01

 exting

0.94

storm

0.94

storms

0.88

trap

0.87

brand

0.86

starter

0.84

fly

0.81

plug

0.81

Activations Density 2.279%

terms related to fire incidents

No Comments

No Known Activations