INDEX

Explanations

comparative terms and indications of severity

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

ocene

-0.62

rieg

-0.60

#$

-0.58

dinand

-0.58

iste

-0.58

 Scrib

-0.58

DEM

-0.57

olan

-0.55

oplan

-0.54

POSITIVE LOGITS

 importantly

0.95

versely

0.85

 guessed

0.78

 interestingly

0.78

 furthermore

0.76

 incidentally

0.74

 irony

0.72

 aside

0.71

 moreover

0.71

FY

0.69

Activations Density 1.927%

comparative terms and indications of severity

No Comments

No Known Activations