INDEX

Explanations

names of individuals and organizations mentioned in news 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_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

 acron

-0.73

 sofa

-0.69

 operating

-0.69

 recomb

-0.68

 supers

-0.67

TRA

-0.67

 roadside

-0.66

COM

-0.66

 couch

-0.66

 shuttle

-0.66

POSITIVE LOGITS

ason

1.15

reau

1.10

ault

1.09

chuk

1.08

ovich

1.08

alez

1.07

orst

1.06

ardi

1.05

neau

1.03

ÃŃn

1.03

Activations Density 0.132%

names of individuals and organizations mentioned in news articles

No Comments

No Known Activations