INDEX

Explanations

phrases related to technical details and procedures

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

eur

-0.75

eering

-0.74

apo

-0.69

urally

-0.67

paio

-0.66

gregation

-0.66

Wan

-0.63

aldo

-0.62

ILLE

-0.61

ä½ľ

-0.61

POSITIVE LOGITS

kick

1.09

down

1.04

downs

0.92

outs

0.87

drop

0.81

backs

0.79

phrine

0.79

plet

0.78

 hints

0.78

lights

0.74

Activations Density 0.020%

phrases related to technical details and procedures

No Comments

No Known Activations