INDEX

Explanations

mentions of mortality or situations where something is no longer functioning or useful

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

orney

-0.67

 CLSID

-0.66

ECA

-0.65

Wiz

-0.64

zzle

-0.60

 Cout

-0.60

zac

-0.59

 Brands

-0.59

é¾įåĸļå£«

-0.59

%]

-0.58

POSITIVE LOGITS

spin

1.17

locked

1.13

liest

1.06

beat

1.02

weight

0.97

 reckoning

0.91

locks

0.89

 carc

0.87

lier

0.86

heading

0.86

Activations Density 0.545%

mentions of mortality or situations where something is no longer functioning or useful

No Comments

No Known Activations