INDEX

Explanations

mentions of specific products or details regarding products, possibly related to reviews or descriptions

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

-0.84

 newsp

-0.69

catentry

-0.67

igible

-0.64

 initials

-0.63

ledged

-0.62

 unpublished

-0.61

 Caldwell

-0.60

isen

-0.59

POSITIVE LOGITS

 Connection

0.89

 Therapy

0.89

 Consumption

0.89

 Accessories

0.85

 Products

0.84

 Removal

0.83

 Matters

0.83

 Digest

0.82

 Supplement

0.80

 Donation

0.79

Activations Density 0.184%

mentions of specific products or details regarding products, possibly related to reviews or descriptions

No Comments

No Known Activations