INDEX

Explanations

proper nouns related to news articles and stories

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

 disarm

-0.71

 academ

-0.71

 campuses

-0.69

aples

-0.69

 resolutions

-0.66

 skelet

-0.66

 warr

-0.65

 inclined

-0.64

 collisions

-0.63

 unimaginable

-0.63

POSITIVE LOGITS

*.

0.91

Äĵ

0.83

Åį

0.82

 Kare

0.79

-.

0.78

qa

0.76

Ä

0.75

_.

0.74

*,

0.74

XXX

0.73

Activations Density 2.213%

proper nouns related to news articles and stories

No Comments

No Known Activations