INDEX

Explanations

references to specific terms like product names, company names, and technical details

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

Ö¼

-0.92

enegger

-0.84

alities

-0.83

puter

-0.83

ĸļ

-0.75

ROR

-0.74

ites

-0.73

rine

-0.73

 Jagu

-0.72

aiman

-0.70

POSITIVE LOGITS

 Centauri

1.53

volent

0.95

cephal

0.83

Bay

0.81

 Quad

0.80

Draft

0.76

Phi

0.73

 Males

0.73

HA

0.71

 rays

0.70

Activations Density 0.021%

references to specific terms like product names, company names, and technical details

No Comments

No Known Activations