INDEX

Explanations

phrases related to expressing opinions or preferences

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

 ÂŃ

-0.82

anwhile

-0.61

angering

-0.60

 amid

-0.58

ashington

-0.58

 Hussein

-0.57

 reportedly

-0.57

ospital

-0.57

iberal

-0.56

 seized

-0.56

POSITIVE LOGITS

 alot

0.99

 kinda

0.95

 devs

0.93

:)

0.84

 setups

0.80

 pokemon

0.79

 haha

0.78

;)

0.77

 reddit

0.76

 optimization

0.75

Activations Density 7.741%

phrases related to expressing opinions or preferences

No Comments

No Known Activations