INDEX

Explanations

structured information related to government policies or actions

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

 Scarlett

-0.68

ï¸

-0.65

rish

-0.64

 Niet

-0.62

 Happ

-0.60

 Hazel

-0.59

Khe

-0.59

eva

-0.58

itely

-0.57

Doc

-0.56

POSITIVE LOGITS

 guiActiveUnfocused

1.17

 federation

0.90

 territories

0.88

¬¼

0.85

 feder

0.84

inav

0.82

 organisation

0.80

verning

0.80

wide

0.79

rongh

0.79

Activations Density 2.260%

structured information related to government policies or actions

No Comments

No Known Activations