INDEX

Explanations

phrases related to entirety or completeness

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

KER

-0.72

chairs

-0.66

tips

-0.65

none

-0.63

bid

-0.62

claimed

-0.62

friends

-0.61

writers

-0.60

jin

-0.60

immers

-0.60

POSITIVE LOGITS

heartedly

1.40

hearted

1.22

 ordeal

0.99

gam

0.91

 thing

0.88

 spectrum

0.88

 ecosystem

0.78

 globe

0.78

 affair

0.76

 swath

0.76

Activations Density 0.960%

phrases related to entirety or completeness

No Comments

No Known Activations