INDEX

Explanations

phrases related to specific tasks or steps within instructions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iola

-0.80

uay

-0.77

hap

-0.76

ubb

-0.76

ulz

-0.75

ordable

-0.73

anta

-0.73

leased

-0.71

uga

-0.68

ificial

-0.67

POSITIVE LOGITS

 ratio

1.28

 ratios

1.01

 trope

0.96

 initiative

0.95

 scenario

0.94

 Ratio

0.94

 clause

0.93

 mantra

0.91

 distinction

0.91

 mentality

0.89

Activations Density 0.550%

phrases related to specific tasks or steps within instructions

No Comments

No Known Activations