INDEX

Explanations

different iterations of the name "Iro"

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

ervative

-0.93

iple

-0.86

itect

-0.80

ennett

-0.74

ŃĶ

-0.72

alyst

-0.71

retion

-0.69

arians

-0.69

ulhu

-0.68

alez

-0.68

POSITIVE LOGITS

zzi

1.16

phant

1.15

tto

1.08

tti

1.05

tes

0.97

ku

0.90

zes

0.90

tic

0.88

tel

0.85

tt

0.84

Activations Density 0.017%

different iterations of the name "Iro"

No Comments

No Known Activations