INDEX

Explanations

numerical references such as quantities or steps in a process

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

actionDate

-0.65

 sacked

-0.64

 Jagu

-0.63

 swearing

-0.61

rolet

-0.60

Fas

-0.58

 Exodus

-0.56

 slogan

-0.56

 Sacrament

-0.56

 folly

-0.56

POSITIVE LOGITS

nd

1.20

Â½

1.11

rd

1.10

WD

1.05

0.96

ÃĹ

0.94

ÏĢ

0.91

yrs

0.90

%-

0.89

âĺħ

0.89

Activations Density 0.140%

numerical references such as quantities or steps in a process

No Comments

No Known Activations