INDEX

Explanations

review and analysis of individuals and their actions

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

ĸļ

-0.80

 guiActiveUn

-0.76

ceivable

-0.75

igham

-0.74

erey

-0.66

atars

-0.66

ulhu

-0.63

 familiar

-0.63

ModLoader

-0.62

rontal

-0.61

POSITIVE LOGITS

nered

0.82

 luck

0.77

 outweigh

0.72

 Practices

0.69

 outcomes

0.68

smanship

0.68

luck

0.68

 outwe

0.67

ounters

0.66

 mate

0.66

Activations Density 2.249%

review and analysis of individuals and their actions

No Comments

No Known Activations