INDEX

Explanations

medical and health-related terms and concepts

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

undai

-0.73

aples

-0.68

vae

-0.62

idden

-0.61

aband

-0.59

 Towers

-0.59

 drafts

-0.58

 annex

-0.58

ãģį

-0.57

endars

-0.57

POSITIVE LOGITS

naires

1.00

 whereby

0.97

lessness

0.87

of

0.85

naire

0.84

iest

0.80

iness

0.77

 anew

0.71

ophys

0.71

iasis

0.70

Activations Density 1.757%

medical and health-related terms and concepts

No Comments

No Known Activations