INDEX

Explanations

words related to computer programming and instructions

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

arians

-0.76

ually

-0.69

 Catalyst

-0.68

 Coulter

-0.68

UAL

-0.67

alam

-0.67

 eminent

-0.64

 signific

-0.64

arella

-0.63

 Whites

-0.62

POSITIVE LOGITS

cks

1.11

bably

1.11

dden

1.08

spective

1.08

ving

1.04

dy

1.03

phy

1.03

ffe

1.02

ppo

1.01

dding

1.01

Activations Density 0.788%

words related to computer programming and instructions

No Comments

No Known Activations