INDEX

Explanations

phrases related to large scale entities or concepts, particularly focusing on the term "big."

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

Dialogue

-0.77

istry

-0.68

manship

-0.67

 Michaels

-0.66

roma

-0.66

 Emin

-0.66

 Nich

-0.65

ICLE

-0.65

onto

-0.65

chron

-0.63

POSITIVE LOGITS

gest

1.10

oted

1.07

 chunk

0.99

 chunks

0.98

 swath

0.97

 intestine

0.97

 scale

0.97

 quantities

0.93

 swat

0.91

 amounts

0.90

Activations Density 2.919%

phrases related to large scale entities or concepts, particularly focusing on the term "big."

No Comments

No Known Activations