INDEX

Explanations

phrases related to specific technical or medical procedures or protocols

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

dozen

-0.69

wcs

-0.67

Ney

-0.63

rez

-0.62

bestos

-0.61

 Thiel

-0.60

ipl

-0.60

efully

-0.60

Jimmy

-0.58

 Jimmy

-0.58

POSITIVE LOGITS

 itself

0.94

 Productions

0.82

's

0.80

âĦ¢:

0.77

 herself

0.72

 Presents

0.72

 Membership

0.71

Â®

0.70

 Implementation

0.69

prototype

0.68

Activations Density 1.988%

phrases related to specific technical or medical procedures or protocols

No Comments

No Known Activations