INDEX

Explanations

information related to spatial orientation and direction, specifically involving movement and alignment

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

egu

-0.73

displayText

-0.69

ufact

-0.68

gdala

-0.68

anan

-0.66

unts

-0.66

tained

-0.65

inez

-0.65

soDeliveryDate

-0.65

TIT

-0.64

POSITIVE LOGITS

 axis

1.05

 pend

1.03

 downward

1.01

 symmetry

1.00

 downwards

0.98

 direction

0.98

 motion

0.93

 jerk

0.93

 shif

0.92

 rotating

0.90

Activations Density 1.680%

information related to spatial orientation and direction, specifically involving movement and alignment

No Comments

No Known Activations