INDEX

Explanations

references to responses, reactions, or replies

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

 thereto

-0.74

 thereof

-0.69

 Ribbon

-0.68

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.68

 trans

-0.67

 Cerberus

-0.66

 âĢ¦"

-0.66

(),

-0.65

".

-0.65

 formerly

-0.64

POSITIVE LOGITS

resa

1.24

roximately

1.16

xiety

1.09

jamin

1.09

vertising

1.08

withstanding

1.06

odore

1.06

ropolitan

1.02

alyst

0.99

icester

0.97

Activations Density 1.062%

references to responses, reactions, or replies

No Comments

No Known Activations