INDEX

Explanations

phrases related to qualifications and expertise

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

izards

-0.75

IUM

-0.74

SHIP

-0.69

YS

-0.69

puters

-0.68

Deal

-0.67

wagen

-0.67

ITED

-0.65

stals

-0.63

ORPG

-0.63

POSITIVE LOGITS

 foremost

1.19

 brightest

0.85

 third

0.84

 simplest

0.81

 longest

0.81

liest

0.81

 largest

0.81

 oldest

0.78

rogens

0.77

 poorest

0.77

Activations Density 0.062%

phrases related to qualifications and expertise

No Comments

No Known Activations