INDEX

Explanations

phrases related to the beginning or end of texts or articles

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

entric

-0.72

enaries

-0.68

viz

-0.66

inational

-0.66

ities

-0.63

uary

-0.62

ision

-0.59

ittees

-0.59

Osc

-0.58

ittee

-0.57

POSITIVE LOGITS

packs

1.30

stab

1.22

dated

1.20

GROUND

1.17

packing

1.14

tracking

1.12

gam

1.07

lash

1.05

doors

1.04

haul

1.02

Activations Density 0.293%

phrases related to the beginning or end of texts or articles

No Comments

No Known Activations