INDEX

Explanations

mentions or references to laboratories or lab-related activities

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_fs6144-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_6144

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

6,144

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

atra

-0.70

 Marketable

-0.65

theless

-0.65

isms

-0.64

 Ceres

-0.63

pend

-0.63

lihood

-0.61

ACTED

-0.58

cut

-0.57

 constitu

-0.57

POSITIVE LOGITS

yrinth

1.38

rador

1.36

elling

1.24

rosse

1.07

orer

1.05

elled

1.04

orers

1.01

orously

0.92

eled

0.91

rade

0.89

Activations Density 0.261%

mentions or references to laboratories or lab-related activities

No Comments

No Known Activations