INDEX

Explanations

phrases related to equality and fairness

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_fs6144-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_6144

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

6,144

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

iatric

-0.69

lov

-0.67

OLOG

-0.67

 Jere

-0.67

olog

-0.64

clips

-0.64

hood

-0.64

soever

-0.63

Nazi

-0.61

Jere

-0.61

POSITIVE LOGITS

 balance

0.99

balance

0.96

balanced

0.95

amental

0.94

 balancing

0.93

 between

0.93

 imbalance

0.91

 balanced

0.89

 sheet

0.83

 balances

0.83

Activations Density 0.285%

phrases related to equality and fairness

No Comments

No Known Activations