INDEX

Explanations

words related to regulations and procedures

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

lihood

-0.90

IBLE

-0.80

UAL

-0.77

sie

-0.69

SOS

-0.68

ãģį

-0.67

 Instruments

-0.66

VERTIS

-0.66

 Scarlet

-0.65

ATURES

-0.65

POSITIVE LOGITS

itone

1.31

celona

1.27

becue

1.24

bara

1.15

iatric

1.12

bed

1.10

riers

1.10

bell

1.08

bers

1.07

rier

1.03

Activations Density 0.611%

words related to regulations and procedures

No Comments

No Known Activations