INDEX

Explanations

terms related to data analysis, processing, and extraction

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

çīĪ

-0.89

Rail

-0.82

FactoryReloaded

-0.76

CLASSIFIED

-0.74

osate

-0.74

soDeliveryDate

-0.72

MpServer

-0.70

çĶ°

-0.69

lez

-0.68

icum

-0.68

POSITIVE LOGITS

 ALWAYS

0.97

 imperfect

0.95

 unpredict

0.92

 mistakes

0.89

 finite

0.88

 inevitably

0.85

 losers

0.83

 always

0.83

 abstract

0.82

 bounces

0.81

Activations Density 3.500%

terms related to data analysis, processing, and extraction

No Comments

No Known Activations