INDEX

Explanations

instances of the word "working" in various contexts

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

antha

-0.76

ylon

-0.71

 Augustus

-0.66

emonic

-0.66

 Bubble

-0.66

anamo

-0.64

 constitu

-0.63

xual

-0.62

angular

-0.61

iren

-0.60

POSITIVE LOGITS

 tirelessly

1.13

bench

1.08

hops

1.03

 diligently

1.03

 collabor

0.99

 ethic

0.98

 overtime

0.97

heet

0.94

paces

0.86

station

0.85

Activations Density 0.333%

instances of the word "working" in various contexts

No Comments

No Known Activations