INDEX

Explanations

event details or schedules

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dictionary

-0.61

 trademarks

-0.59

 Wiki

-0.57

 trademark

-0.55

volent

-0.55

 Polit

-0.54

 Clever

-0.53

lihood

-0.53

 whistleblower

-0.53

 Unicode

-0.53

POSITIVE LOGITS

aturdays

1.05

nesday

0.79

TBA

0.75

aturday

0.73

dule

0.73

 thereafter

0.72

noon

0.72

interrupted

0.71

clock

0.70

 rehearsal

0.69

Activations Density 0.902%

event details or schedules

No Comments

No Known Activations