INDEX

Explanations

words related to habits or routines

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

meal

-0.71

 Dawkins

-0.70

 skirts

-0.64

 Dempsey

-0.64

chnology

-0.63

imates

-0.62

CLS

-0.61

 Ferdinand

-0.59

 thickness

-0.59

 underest

-0.58

POSITIVE LOGITS

ilitation

1.57

itual

1.40

itable

1.29

ilit

1.23

itation

1.21

itant

1.19

ilitating

1.19

ited

1.11

itating

1.10

itability

1.10

Activations Density 0.017%

words related to habits or routines

No Comments

No Known Activations