INDEX

Explanations

references to limited or exclusive experiences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 renegoti

-0.82

 litigation

-0.78

 administr

-0.75

 arbitration

-0.75

 bipolar

-0.70

 involuntary

-0.70

 counseling

-0.70

 transact

-0.69

 wrongful

-0.69

 nonviolent

-0.69

POSITIVE LOGITS

 Featuring

1.24

eaturing

1.03

 Beaut

0.93

 Located

0.91

Located

0.87

 Thankfully

0.86

Fans

0.85

 adorned

0.80

 Sadly

0.79

 sporting

0.79

Activations Density 0.505%

references to limited or exclusive experiences

No Comments

No Known Activations