INDEX

Explanations

references to specific locations in the context of governmental activities

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

 indicators

-0.60

 resemblance

-0.59

FontSize

-0.58

Trend

-0.57

%"

-0.56

 disagreement

-0.56

OWS

-0.56

Rating

-0.55

 Nielsen

-0.54

pse

-0.54

POSITIVE LOGITS

 destinations

0.89

via

0.89

 shores

0.83

 undet

0.79

 beck

0.78

 whence

0.73

 airst

0.72

 fray

0.69

agar

0.69

 unaccompanied

0.69

Activations Density 0.251%

references to specific locations in the context of governmental activities

No Comments

No Known Activations