INDEX

Explanations

names or proper nouns related to various individuals

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

=-=-=-=-

-0.71

ptives

-0.69

xon

-0.69

Wan

-0.68

utics

-0.67

ITE

-0.66

RECT

-0.65

coni

-0.65

âĸĳâĸĳ

-0.65

USS

-0.64

POSITIVE LOGITS

reau

1.09

oun

0.96

iquette

0.93

mond

0.88

emouth

0.87

lers

0.86

spe

0.86

NESS

0.85

ness

0.84

eh

0.83

Activations Density 0.019%

names or proper nouns related to various individuals

No Comments

No Known Activations