INDEX

Explanations

legal and academic terminology and emotional expressions related to intense negative feelings

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.73

 Nare

-0.72

 IMAGES

-0.70

amins

-0.69

arnaev

-0.68

 towed

-0.64

 portions

-0.62

 raided

-0.62

 cleaned

-0.62

 Shuttle

-0.62

POSITIVE LOGITS

ness

1.22

eness

1.19

activity

1.17

 insanity

1.17

 greatness

1.11

itism

1.11

 heroism

1.10

ility

1.10

ity

1.09

 blindness

1.08

Activations Density 0.549%

legal and academic terminology and emotional expressions related to intense negative feelings

No Comments

No Known Activations