INDEX

Explanations

phrases related to the presence or absence of specific elements or entities within a context

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

imb

-0.83

jab

-0.74

geon

-0.73

ousse

-0.71

jug

-0.70

bard

-0.67

 fixme

-0.67

cipled

-0.67

¥ŀ

-0.66

romy

-0.66

POSITIVE LOGITS

uated

0.95

 presence

0.83

enance

0.83

lihood

0.78

uate

0.76

 Presence

0.73

quo

0.72

uality

0.71

uating

0.71

iveness

0.71

Activations Density 0.114%

phrases related to the presence or absence of specific elements or entities within a context

No Comments

No Known Activations