INDEX

Explanations

the word "all"

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

 Patriarch

-0.67

 Chancellor

-0.63

pu

-0.58

 chancellor

-0.57

 proverb

-0.57

 Koen

-0.56

jun

-0.55

SI

-0.55

essor

-0.55

lav

-0.54

POSITIVE LOGITS

 whatsoever

0.83

 except

0.83

 else

0.79

ocating

0.77

ãĥīãĥ©ãĤ´ãĥ³

0.76

onse

0.72

 sidx

0.71

 anymore

0.69

ogene

0.68

 conceivable

0.67

Activations Density 0.016%

the word "all"

No Comments

No Known Activations