INDEX

Explanations

phrases related to groupings or sets of people

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

together

-0.71

Ranked

-0.70

BACK

-0.70

wcsstore

-0.69

Reply

-0.67

je

-0.67

/"

-0.66

="#

-0.66

backs

-0.66

estones

-0.65

POSITIVE LOGITS

 unsuspecting

0.88

 passers

0.85

 whoever

0.80

 those

0.79

 disgruntled

0.78

beh

0.78

 mortals

0.78

 regulators

0.77

 anyone

0.74

 onlook

0.74

Activations Density 0.113%

phrases related to groupings or sets of people

No Comments

No Known Activations