INDEX

Explanations

mentions of chronological rankings or the order of events

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

ledge

-0.81

utics

-0.79

allah

-0.78

today

-0.78

md

-0.77

each

-0.76

erved

-0.76

dain

-0.76

their

-0.72

rs

-0.72

POSITIVE LOGITS

 installment

1.24

 iteration

1.10

 thing

1.08

 batch

1.05

 incarnation

1.04

 straw

1.03

 edition

1.01

 version

0.96

 frontier

0.96

 piece

0.96

Activations Density 2.941%

mentions of chronological rankings or the order of events

No Comments

No Known Activations