INDEX

Explanations

specific locations or places

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

apult

-0.77

termin

-0.68

ME

-0.64

-0.62

TAG

-0.60

rolet

-0.58

rieve

-0.58

 Anyway

-0.57

fml

-0.57

Translation

-0.57

POSITIVE LOGITS

upon

1.66

soever

1.26

abouts

1.07

 they

0.88

fore

0.83

 temperatures

0.73

he

0.72

we

0.71

she

0.68

 homosexuality

0.67

Activations Density 0.374%

specific locations or places

No Comments

No Known Activations