INDEX

Explanations

examples of proper nouns, potentially related to news or research articles

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EEP

-0.67

eanor

-0.66

 sorely

-0.66

 ultrasound

-0.65

Goo

-0.62

doms

-0.62

ily

-0.62

uckland

-0.61

 unregulated

-0.60

 nomine

-0.59

POSITIVE LOGITS

colo

1.25

asso

1.12

ross

1.10

nic

1.07

chu

0.93

alog

0.88

aret

0.88

asus

0.86

atin

0.85

Pic

0.85

Activations Density 0.030%

examples of proper nouns, potentially related to news or research articles

No Comments

No Known Activations