INDEX

Explanations

technical and scientific terms, particularly related to computer technology and components

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

ĺħ

-0.67

ãĤ´ãĥ³

-0.66

 teasp

-0.65

CHAT

-0.64

xtap

-0.62

 entirety

-0.59

ĳå£«

-0.59

ï¸

-0.57

¬¼

-0.57

 Beaut

-0.57

POSITIVE LOGITS

 than

2.11

than

1.97

 Than

1.47

erous

0.90

efficient

0.84

worldly

0.68

ultz

0.67

rab

0.66

 compared

0.64

TH

0.63

Activations Density 1.649%

technical and scientific terms, particularly related to computer technology and components

No Comments

No Known Activations