INDEX

Explanations

the word "core" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ufact

-0.77

hops

-0.72

 Laksh

-0.71

ilon

-0.69

chedel

-0.69

TN

-0.67

pload

-0.64

uthor

-0.64

ibal

-0.64

xiety

-0.63

POSITIVE LOGITS

 tenets

1.01

stones

0.85

 essence

0.85

stay

0.85

stone

0.85

opsis

0.83

 pillars

0.81

 pillar

0.80

lig

0.80

 principles

0.80

Activations Density 0.021%

the word "core" in various contexts

No Comments

No Known Activations