INDEX

Explanations

phrases or words related to approaching or being approached

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

pite

-0.79

cell

-0.75

glass

-0.72

marine

-0.68

rats

-0.67

fre

-0.66

inker

-0.66

runs

-0.66

boys

-0.65

bel

-0.64

POSITIVE LOGITS

 Ceres

0.81

 Lans

0.70

atown

0.69

IDENT

0.69

 Thumbnails

0.67

adata

0.66

 intersections

0.66

 adulthood

0.66

 puberty

0.65

ADRA

0.65

Activations Density 0.021%

phrases or words related to approaching or being approached

No Comments

No Known Activations