INDEX

Explanations

names of various individuals

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

Negative Logits

 millennials

-0.60

 visitation

-0.56

giving

-0.55

 Millenn

-0.55

 Aval

-0.55

 Hemisphere

-0.54

 trad

-0.53

 Millennials

-0.53

 reversible

-0.53

 mediation

-0.53

POSITIVE LOGITS

urger

0.90

zyk

0.90

augh

0.89

yk

0.84

urn

0.82

acher

0.82

ansky

0.82

owicz

0.81

lich

0.81

orst

0.81

Activations Density 0.101%

names of various individuals

No Known Activations