INDEX

Explanations

references to events, associations, amusements, and preferences concerning numbers and lists

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

hers

-0.79

cles

-0.74

 gifted

-0.73

gewater

-0.67

reach

-0.67

nesday

-0.64

undai

-0.64

 neighb

-0.63

uve

-0.63

elled

-0.62

POSITIVE LOGITS

0.94

0.92

rpm

0.92

%-

0.91

ances

0.81

 minutes

0.80

ACP

0.80

0.79

0.76

%:

0.76

Activations Density 0.057%

references to events, associations, amusements, and preferences concerning numbers and lists

No Comments

No Known Activations