INDEX

Explanations

entities or proper names that are associated with specific locations or cultural references

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

drm

-0.73

ratulations

-0.69

yip

-0.64

hander

-0.62

 Compatibility

-0.60

rish

-0.58

 Scythe

-0.58

kefeller

-0.58

ufact

-0.57

usercontent

-0.57

POSITIVE LOGITS

igham

0.82

 dots

0.78

opolis

0.75

Ö¼

0.72

 Sole

0.70

oult

0.68

cially

0.67

adr

0.67

Ã«

0.66

auga

0.65

Activations Density 0.435%

entities or proper names that are associated with specific locations or cultural references

No Comments

No Known Activations